Carving Out a Niche: Making the Reading Room Globally Available for the First Time
Jamie Corson (firstname.lastname@example.org)
Library Assistant II, Collection Management
The University of Memphis, University Libraries
Brigitte Billeaudeaux (email@example.com)
Library Assistant II, Preservation and Special Collections
The University of Memphis, University Libraries
Originally presented at the Tennessee Library Association Annual Conference (Chattanooga, TN) in April 2013.
Conference Abstract: This session presents two institutions seeking to create a digital presence for unique material that has been underrepresented in academic research. Utilizing collaborative efforts, they are making localized collections globally available for the first time to scholars and professionals who have relied on private collections and hobbyists in the past.
Digital collections are becoming increasingly common in libraries and archives, and as a result, our users are expecting to find more and more content online. In addition to meeting users’ expectations, digital technologies offer librarians and archivists new tools for preserving the historical record and ensuring future access to documents of enduring value. Digital collections allow libraries to reach a broader audience than ever before. This is especially true for special collections material. Historically, special collections departments have been full of hidden collections that are not easily accessible and are often not sufficiently described. Interested researchers are frequently told to speak with the curator and are occasionally required to explain their research and schedule a specific time to consult collections. The process can be confusing or intimidating, especially to undergraduates or novice researchers. Going digital can change that perspective and introduce special collections to a whole new audience.
Through the creation of digital collections, materials that were once relegated to the reading room are now available worldwide with the click of button. This is not to say that people will no longer utilize physical collections, but having a digital presence gives users an idea of what you have and helps them make more informed decisions concerning additional research. How can libraries embrace the new digital landscape and make the most of their collections? This paper provides examples of how different institutions have utilized collaborative efforts to get projects off the ground. We will discuss our experiences and detail the types of issues that may arise when creating digital collections for the first time.
The André Studios Collection
-- Jamie Corson, Library Assistant II, University of Memphis Libraries, Collection Management Department / Previously: Project Intern, Fashion Institute of Technology, State University of New York
Before I moved to Tennessee, I lived in New York City and worked within the Special Collections Department at the Fashion Institute of Technology’s Gladys Marcus Library. While there, I primarily worked on the André Studios project, a digitization effort involving over 4,500 fashion sketches and drawings from 1930-1941. André Studios was a New York sketch firm founded in 1930 by Pearl Levy Alexander and her business partner Leonard Schwartzbach. The subscription-based service employed sketch artists to attend couture fashion shows, usually in Paris, and copy the original fashions from leading designers. The sketches would then be taken back to New York and reworked, many times into color drawings, which would then be compiled into subscription catalogues. When the company closed its doors in the 1970s, the collection of sketches was separated and donated in unequal amounts to three New York City institutions: New York Public Library (NYPL), The Fashion Institute of Technology (FIT), and Parsons New School for Design (André Studios, n.d.). The André Studios project was a grant-funded collaboration between FIT and NYPL, with the primary goal of reuniting the André Collection and making the material available to the public in a digital format. The secondary objective in digitizing this material was to develop a prototype for a web portal catering to fashion researchers.
Addressing Preservation Needs
FIT’s André Collection had not previously been processed, and each sketch was still housed in the original binding. The binders were organized by season and bound using different techniques. Nails held the early material together, while the later books were bound using rope and glue. In order to digitize the individual images, each sketch had to be carefully removed from the binding. I was able to gently pull the nails from the board and remove the fragile paper without much damage; however, the sketches bound by rope and glue were more difficult to remove. In order to ensure that the sketches would not be torn, I carefully cut each sketch along the existing holes to remove it from the binding. Each sketch was then re-housed loosely in acid free folders and transferred to boxes for easy storage and transportation. Some of the earlier sketches were particularly fragile and were encased in Mylar sleeves for added protection. Each folder was labeled with ID numbers corresponding to the set of sketches within and included data that indicated where the folder was physically located (ex. Folder 1, Box 1).
Prior to the project, FIT had a brief description for the André Collection, which was findable through the OPAC; however, like most archival material, the collection was not processed at the item level. We knew that we would be using Dublin Core to describe each image, but we also needed to create preliminary descriptors for each sketch before it could be sent off for scanning. Because FIT did not have access to NYPL’s digital processing system (Hades), data was input into Excel spreadsheets, which could then be ingested into NYPL’s system and assigned to the appropriate digitized image.
Because the individual sketches had never been described, we faced questions concerning what type of information needed to be captured. For example, what would an appropriate title look like?
The preliminary metadata included the following seven data fields:
Root ID: corresponds to the collection as a whole
Parent ID: corresponds to the book of sketches (i.e. Winter Sketches 1939)
SIB Sequence: corresponds to the page within each volume
Title: title given to each individual sketch
Indicator bytes: number of spaces to skip in order to reach relevant information
Page Number: refers to the number written on each sketch by the creator
Classmark: FIT identifier number written on each sketch
The only subjective information entered was the title. Not being a fashion expert myself, I created a formula whereby I identified the garment (i.e., coat) and a couple of distinguishing features (i.e., coat with brown fur collar and red buttons). If the sketch referenced the name of a particular designer, I also included that information in the title. Part of my goal in creating the descriptive titles was to ensure that those digitizing the sketches would feel comfortable confirming that that data matched the item in hand. Each sketch contained many additional descriptors that would not fit into the supplied title, but once the sketches were scanned, we could then go back and supply more extensive metadata using Dublin Core as the standard.
One of the benefits to supplying item specific metadata was that we learned more about the collection and the function of André Studios. Many of the books included code sheets that listed the names of designers along with a corresponding alphabetic code. The codes were then included on subsequent sketches as a way of secretly identifying which designers were being copied. In these cases, I was able to include the name of the original designer in the sketch’s title. Users can now search the collection by name and receive a list of fashions that were copied from a particular designer. The existence of the code sheets was not known prior to the digitization project, and searching the physical collection in this manner would have been extremely time consuming.
Because the project was a collaborative effort, we ran into many unforeseen issues when attempting to reconcile the practices of different institutions. FIT did not have the resources to digitize material on a large scale, nor did they have sufficient server space devoted to Special Collections. NYPL, as project partner, was responsible for digitizing and housing the images on their own servers. Most of the problems arose from having to transfer the data created by FIT into NYPL’s repository. The data had to be entered into Excel spreadsheets, then transferred into Hades, and then converted back to Excel spreadsheets. This process required the data to be handled on three occasions as opposed to one. Human error and digital glitches caused several of the numbers to become transposed, which resulted in items being misnumbered. The misnumbering threw off the sequence of titles in such a way that the descriptive titles no longer matched the corresponding sketches. We feared that someone would need to manually correct hundreds of records; however, it turned out that the error occurred when exporting the data from Hades back into the Excel spreadsheet. The metadata was correct in the database; therefore, we were able to run a subsequent export, and the spreadsheets were corrected. The process of identifying the error and locating the origin of the problem cost precious staff time. In future collaborations, it would be wise to limit the handling of records, if possible. The potential for error increases with the number of individuals and processes involved in data manipulation.
We also ran into the unexpected problem of having to reconcile Dublin Core standards with the data transfer process. The repeated metadata fields (i.e., ‘Subject’) were not recognized when transferring the data from Excel spreadsheets to Hades, causing the information to be lost. At the time, there were three possible solutions, each involving extra time or effort on the part of one of the partners. In the end, we relied on a combination of data programming and extra data entry work to ensure all of the information was input correctly into the management system.
The third issue that required collaboration involved the digitization of the sketches themselves. Although the process was not done on site, we were able to visit NYPL’s Digital Imaging Unit in Long Island City, where we had the opportunity to view the scanners and digital cameras being used for the project. We brought a colored sketch to the facility in order to observe the process, and after the sketch was photographed, we noticed the digital rendering was not completely true to the color of the original. Color is an especially important component of the sketches and serves as valuable information to fashion researchers. Had the digitization been done onsite, we would have caught the discrepancy immediately. However, even after adjusting the lighting and altering the image using editing tools, the color was not identical to that of the original. There would be no way to spend the amount of time required to get each color exactly right when considering there are over 4,000 sketches in the collection. The color would have to be sacrificed for completing the project in a timely manner.
Designing a Web Portal
One of the benefits of collaboration is that each institution is able to play to their strengths. NYPL and their web development team did the actual construction of the web platform, which was built in-house. Because FIT had more experience with the needs of fashion researchers, we took the lead in designing the site by creating wireframes and navigational flowcharts with the users’ needs in mind. Fashion researchers come from a variety of fields. In addition to students, FIT receives requests from visiting scholars, authors, film researchers, fashion professionals, and television production companies. More than other disciplines, fashion research is dependent on images. While textual material can be crucial for filling in gaps of information, the visual material is what researchers are drawn to first. In speaking to graduate students who frequent FIT’s Special Collections, we learned that the websites they repeatedly used were those that provided visual material in an easy to use format. The data should be helpful in facilitating the search results, but should not hinder access to the image since that is what researchers are there to see. The site’s design should highlight the material and exhibit simple text and vibrant imagery. We displayed one of the digitized sketches prominently on the homepage and ensured that each visit to the page generated a different drawing. The hope was that in seeing the image, users would be enticed to search (or browse) the collection to find additional material.
Most researchers are looking for materials from a specific designer or time period. While a garment’s individual features are important, they are not the initial qualifier behind most search requests. This was a key factor in determining how we designed the search and browse features of the web platform. Users can browse the collection by clicking one of the many suggested search terms or tags. Once a set of results is compiled, one has the option to narrow the results using a series of filters. It was also important to us that a keyword search be prominently displayed on the homepage since most users will initially search using a keyword. Keyword searching is also beneficial in capturing information that may be entered in various places--title, notes, subject, etc. After observing the search methods favored by fashion researchers, it became evident that most start with a broad search topic and then subsequently become more focused. Because of this practice, we designed a method of search in which a user has the ability to search for a keyword and then use filters to get exactly what is needed. For example, a researcher may be looking for fashion from the designer Mainbocher. After performing a keyword search and receiving many pages of results, the user would then have the opportunity to narrow the search by choosing a particular type of garment (i.e., suit) or a specific year. The filter process allows users the capability to cater the search to their specific research needs.
Prototype for the Future
As the creator of a potentially expansive portal, we designed the website with future contributors in mind. Although Parsons New School for Design did not have the resources to take part in the initial project, we wanted to set the framework to make future collaboration possible. I researched other collaborative endeavors such as California’s Local History Digital Resource Project and expanded on existing ideas and practices (Levy & Turner, 2008). In order to make contributing to the project as easy as possible, we discussed the creation of a help section including metadata templates as well as statements on best practices and instructions on how we provided our metadata. If we can standardize the metadata, we ultimately make it easier for researchers to find the materials they seek. To help institutions lacking the finances to contribute, we should also include links to grant options and collaborative projects. The more information we can provide, the better our chances will be for successful collaboration.
In addition to a help section for contributors, it would also be beneficial to include a section devoted to assisting users. For example, it would likely be jarring for users to discover standard shifts between images. While this may not be an issue immediately, we should prepare for this possibility once other institutions begin contributing their collections. In order to minimize users’ discomfort, we can provide a general overview of how the consortium works and include tips on how to use the various methods of searching. It would be helpful to include simple training videos or slide presentations for users who may want additional guidance. Another crucial element to increasing users’ search ability would be the inclusion of a glossary for fashion terminology. Most of these ideas have not been implemented to date, but the framework is there. André Studios 1930-1941: Fashion Drawings and Sketcheswent live in 2012, with the full collection of sketches from FIT and NYPL being available in one location for the first time. I am eager to follow the success of the project and am optimistic that with increased interest in digital collections, many other institutions will have the opportunity to contribute similarly unique and important material.
The National African American Photographic Archive
-- Brigitte Billeaudeaux, Library Assistant II, University of Memphis Libraries, Department of Preservation and Special Collections
History of the Project
The National African American Photographic Archive (http://naapa.memphis.edu) was first conceived and developed by local Memphis business executives Henry and Cheri Rudner of Ampro Industries. Ampro Industries was founded in 1947 by Mr. Irvin Lanksy who owned a dry goods store in what is now historic Beale Street in the heart of downtown Memphis, Tennessee. Beale Street has always been associated with the African American community of Memphis. This is where many African American Memphians shopped, worked, and engaged in social activities. Mr. Lanksy named the products “Ampro” to stand for “American Products.” Over sixty years later, Ampro is a nationally recognized brand that focuses on customer needs, the interest of its employees, and the community (Ampro, n.d.).
In early 2007, Mr. Rudner set up a 501(c) (3) organization and acquired a small in house staff to plan and implement the project. It was not long after these steps that it became apparent that he did not have enough resources to get this project off the ground and sought out the services of the University of Memphis Libraries. Mr. Rudner approached the Dean of Libraries, Dr. Sylverna Ford, about taking the project on with limited funding from Ampro to build a digital archive that would celebrate the African American experience through photographs. This opportunity aligned with the University Libraries’ desire to enter into the realm of digitization projects (Danley, 2011). Out of this meeting a partnership was created and in 2010 the University Libraries brought the nonprofit into the University Library system, under its collections, as a separate nonprofit that would take on the responsibility of acquiring, scanning, curating, displaying, and maintaining African American images of cultural significance under the National African American Photographic Archive, also known as NAAPA. As of June 7, 2011, the organization’s mission states, "The NAAPA project creates, maintains, and provides online access to a virtual collection of photographs that documents the lives and history of African Americans."
The National African American Photographic Archive was developed to celebrate and give members of the African American community an opportunity to have a voice in their cultural heritage. In many cases the African American community has been historically underrepresented to the point of being ignored. This is certainly true of the African American community in Memphis and the Mid-South. The NAAPA project looks to build a digital archive of images that allows African Americans to tell their story through their own lenses. To date, the archive has been seeded with materials taken from the Preservation & Special Collections Department at the University of Memphis Libraries.
Using Seeds to Grow the Collection
The Mississippi Valley Collection (MVC) is Preservation and Special Collections' predominant collection. The department is no stranger to providing our collections to patrons wanting to use to them on the global stage. For many years we have provided materials to authors, directors, educators, and scholars for their research. When NAAPA was first conceived it was meant to be a grassroots project seeded with images from community stakeholders and individuals with images that showed African American life and history. In the beginning, however, the University Libraries was not prepared to legally accept such gifts from community partners. Instead, the NAAPA archive would focus on images taken from several of the collections in the MVC.
Many of our collections have only been findable and accessible through traditional avenues of research and librarianship. In many cases gaining access to our collection means making a trip to the archive because many of our materials are not findable in our online catalog. One of the top priorities of NAAPA is to create accessibility in our department’s collection of materials exhibiting African American life and history. Until recently, NAAPA has been seeded with materials from four primary collections: The Robert R. Church Family of Memphis Collection, the Memphis Press Scimitar Morgue File, various ICON Lots (collections of subject specific images), and images from The Commercial Appeal newspaper.
Initial Evaluation and Planning
I joined the staff of Preservation and Special Collections in July 2010. One of my primary duties was to begin working to get NAAPA up and running. Before I was hired by the University Libraries, the department spent a year preparing to receive the software that NAAPA would use to develop its database and website. Shannon Williams, a graduate student, prepared an arsenal of background materials, notes, and recommendations and guidelines for different aspects of the project. These materials acted as an excellent primer for things that had already been done. Williams, however, did not have the opportunity to implement any of the strategies she had formulated because we had been in the process of transferring NAAPA from AmPro to the University of Memphis.
The first thing I noticed about the images that would be included in NAAPA was there was no comprehensive digital organizing system for them. To this point in the department’s history, the staff had been digitizing images for the public for over a half of a decade. The policy was more of a “scan on demand” public service than a system designed for preservation and display. There was a clear need for organization that would reflect the nature of how things were organized within our collection. My first action was to formulate a naming convention that could be useful for all images and manuscript materials found in the MVC.
The second step was to review the standards by which the department had been scanning and saving their images. The majority of the images had been scanned at 150 or 300 pixels per inch (ppi) and saved in JPEG format. There was no stated policy for scanning and housing images into an archive for the department. The recommendations set forth in the policy and procedures manuals talked extensively about the different formats for the images and made recommendations about which formats to use for storage of digital images, but only gave ranges at in which to scan images. The manuals failed to make a concrete recommendation based on the best practices. Additionally, there was no dedicated space for items from the MVC to live once they had gone through intake and processing. There was a clear need for a baseline archival plan to get the archive started.
With initial needs identified, I got to work on setting up a naming convention that could be translated to all items in the MVC based on their physical location and the collection to which they belong. My department worked closely with the Libraries' systems department who provided us with server space specifically for NAAPA. I came up with a workflow that included scanning, processing, and saving images specified in a manner that organized the images by file format type. These steps lead to the development of training materials for the graduate assistant funded through our project and for others needing a primer on how to operate the scanner and Abode software that we used to build the collection.
Throughout the preparations for the University Libraries to bring NAAPA into its collections, there was some interdepartmental collaboration that took place in setting up guidelines and standards for the project. By the time I joined the project, the NAAPA team decided to use the Dublin Core Metadata Initiative (DCMI) for the project’s metadata, on advice of the cataloging department. This schema was chosen because of its functionality and flexibility. The schema offered basic information and allowed expanded data to be included in image records. This would allow users to have detailed information for using the images outside of the website for personal use with the appropriate permissions.
Later, the NAAPA team decided that instead of using Library of Congress Subject Headings (LCSH) as a controlled vocabulary, to use the Library of Congress’ Thesaurus for Graphic Materials (TGMI). TGMI does a much better job at conveying concepts than subject headings in LCSH. The thesaurus also allows for more geographic flexibility, because many of the terms suggested by the thesaurus can be given geographical notations. In addition to these subject terms, we decided to use Library of Congress Name Authority(LCNAF) records. We also planned to create a local authority file that included both LCNAF names and those not found in the LCNAF.
Issues We Encountered and How We Overcame Them
The Rudners graciously donated the software and licenses to house the images that the project would receive and provide the means to set up an interactive user website for displaying and searching the images. From the outset, we had issues with this software and its server counterpart. Our biggest issue with the product was that it was proprietary. There were no elements of the data or content that went into the database that could be later harvested or migrated without having another type of software produced by the software company. The system was so robust that it was difficult for the library’s staff and the graduate student working on the project to obtain the skills needed to exercise control over the program. Other issues with the software and platform had to do with the system’s interface and its inability to fit into the University of Memphis’ website branding standards.
When we began working on the project we discovered that the donated software was behind three upgrades and to make the program operational we would have to upgrade the software to the current version. Once we upgraded the software, we realized that data and digital content that went into the database could only be exported in a proprietary file format. In order to read or use the content outside of the database program, additional products made by the program creator would have to be purchased. This was undesirable because we knew that one of the institutional goals for NAAPA was to make collections and items findable in the library’s online catalog. In order to do that, the metadata and other collection level data must be interoperable to other systems. This was not possible using the first software.
Another problem we encountered was the level of database building that was required to make databases in the system functional. The level of knowledge needed for database creation went beyond the skillsets of our staff. To date, we have had great success with interdepartmental collaboration on this project, but no University Libraries employees work with NAAPA full time. The original software and platform were robust and could handle as many collections as server space would allow. However, my graduate student and I realized that in order to create a collection within the program, we would have to select and define each field of the metadata that would be used for the collection’s digital content. This was difficult for two novices to master, and it was also time consuming.
The program offered over 150 different elements that could be used in collection level metadata. Very few of those elements were equivalent to the DCMI that we had originally adopted for this project. We chose DCMI because it is flexible and our catalogers were comfortable with walking DCMI metadata into other metadata schemas should this become a future need of the project. The project team decided that in order to make the software’s elements work for the project, we should focus on using the elements that best fit into DCMI and then create a data dictionary that could be used locally to navigate the program’s metadata. Tackling this task would mean taking valuable time away from the collection development of NAAPA.
There were so many steps to take to set up a collection that it became counterproductive for the project. Out of frustration with the issues that we had encountered to date, I was asked to perform an analysis of the time I had spent learning different aspects of this program. We expended over eight thousand dollars in staff time in less than a year in trying to figure out how the program operates. Senior faculty on the project determined that there were better library-centric solutions that were available that could have been pursued with the money and time that had been spent up to this point in the project. The lack of interoperability of the metadata and the inability to take contents stored into the system out of the system contradicted one of our primary goals of making NAAPA available and searchable through the online catalog. The issues that our team encountered with this software would eventually lead us to abandon the use of this product for one much better suited for the needs of an institutional repository.
The last big issue we encountered initially with NAAPA was one of timing. By the time we identified and addressed many of the issues with the software program, almost one academic year had elapsed since work had begun on the project. Many of the plans that had been made to identify and work with community partners to help grow the collection’s holdings had been sacrificed to have a computer program and platform that worked. At the end of the Spring 2011 semester, we had only one collection digitized and we still did not have a fully operational database program to begin building a public site for NAAPA. This was a point of contention for the library’s administration and the NAAPA board because we had set reachable goals to produce 200 images and records within the first year. Obviously, we were unable to reach that goal, given the year we had just spent trying to make our donated software system work. Fortunately for NAAPA and the University Libraries, a big change was about to happen.
Going into year two, our project had around 200 images, a handful of records that we were storing in a spreadsheet, and a “dummy” site that our systems department had created to convey the kinds of things we wanted NAAPA to be. In July 2011 I received word that we were in talks with the provider of our integrated library system about acquiring a content management platform. This content management system would allow us to use DCMI metadata and house images for an unlimited amount of collections. This would allow the University Libraries to expand the scope of the project and go beyond NAAPA to include more materials housed in the library’s collections. In late summer of 2011, the University Libraries adopted Innovative Interface's Content Pro content management system to house NAAPA and to house the University Libraries Digital Repository (ULDR), in which NAAPA was the flagship collection.
There were several benefits to switching to a content management system such as Content Pro. Innovative Interfaces provides better product support than we experienced with our previous system. Innovative allows service tickets to be open, while the previous system required an elaborate GoToMeeting system to serve service needs. A big benefit to using Content Pro is that we are storing our digital content on Innovative's servers, which are located in California. Our library liked this added security feature in regard to preservation needs for the collection. Should on site copies and backups fail or become corrupted, there is always the option obtain a copy of our stored collections from Innovative. Content Pro offers several other attractive features. With this system, we can stream sound and moving image media from offsite servers with few problems. Content Pro allows us to upload multiple images to one record, thus minimizing extensive use of the relation field. Furthermore, Content Pro's default metadata schema is based on DCMI and it allows for expanded Dublin Core elements. Finally, Content Pro has robust searching capabilities, and it allowed us to customize the interface to meet current University of Memphis branding standards.
In the second year of NAAPA we were able to make much greater strides than in the first year that I worked on the project. The project hired a new graduate student for the second year and all University Library employees who would use the content management system, including staff outside the NAAPA project, received Content Pro training. By November 2011, we were well underway with digitizing and populating our first online collection, the Robert R. Church Family of Memphis Collection, and preparing for a platform upgrade that would provide us with the enhanced Content Pro interface that we use today. By the end of the second year we had uploaded and stored over 500 images and records in Content Pro through various collections and had digitized over 900 images that are housed on servers at the University of Memphis Libraries.
At the closing of academic year three and going into year, four we are excelling and moving closer to our initial project goals. We have added the private collection of Earnestine Jenkins, African American art historian and photography expert; we brought our first out of house collections into NAAPA, and we have worked out many of the legal issues that have barred us from pursuing the community components of NAAPA. To date we have over 1,000 images in Content Pro that represent over 900 records. We have decided to start growing NAAPA from within our library family and will be adding capable and interested staff to help us continue to populate existing and new collections in Content Pro. As of its last meeting, our executive board is very pleased with the progress that we have made in the two years that we have been working with NAAPA, and we are currently in talks with board members and outreach faculty about creating curriculum for students and pursuing grant funding to expand the project.
Making the Reading Room Globally Available
NAAPA’s success to date has given us the institutional confidence to move into other areas of preservation and digitization. The department has also partnered with our consortium through LYRASIS to outsource the digitization of many University owned publications, annuals, and oral histories that are housed through the Internet Archive. Making the decision to use Content Pro encouraged the library to expand beyond the scope of NAAPA to highlight collections from different Library collections and activities, such as our branch library in the University’s Music School, the Delta - Everything Southern! conference, and materials from traveling exhibits like the ALA’s Lincoln: The Constitution and the Civil War.
From the very beginning, the University Libraries had a vision with NAAPA. This collection allows our libraries to provide a service on a scale that we have never been able to do before. Making materials like the ones available through the Robert R. Church Family of Memphis Collection or the Multimedia Collection of the Sanitation Workers Strike helps to build on existing scholarship, encourages new areas of study, and allows collection access to anyone who has Internet access. Great strides have been taken to provide end users with as much useful metadata as possible, and the department encourages users to contact us for further information on any digital items As NAAPA continues to grow, we have greater community and education aspirations that will allow this global reading room to teach and empower the people who choose to use and participate in the project.
It has taken a very long time to get to where NAAPA is today. The project has endured many failures and speed bumps, but at the end of it all, we have emerged a stronger collection and a smarter team.
Danley, M. Operational-Level Analysis of Status of NAAPA Project As Of Mid-July 2011.(July 2011). (Draft of opinion based on evidence from professional literature, records of U of M meetings, discussions, etc.)
Levy, T., & Turner, A. (2008). Evolution of a digital collaboration: California's local history digital resource project. Visual Resources Association Bulletin, 35(2) (Summer 2008),94-98.
Fashion Institute of Technology, State University of New York and New York Public Library. (n.d.). André Studios Digitization Project. Retrieved from http://andrestudios.nypl.org/project