Difference between revisions of "Progress Notes"
|Line 1:||Line 1:|
Latest revision as of 11:03, 15 February 2013
This page pertains to "Digitizing the S. D. Cabaniss Papers," an NHPRC-funded grant project.
We had begun digitizing content in December, in order to have material with which to test the work flow, scripts, and display. This gave us a jump on the planned amount of digitization for January. Scripts used, methodology, and work flow were added to the wiki in December and January.
Beginning a bit early enabled Amanda to train one existing Digital Services student, who was available to work the first week of January, while the other did not start until January 11th, and then had to undergo training. The award letter did not arrive until late in January, and the pre-award approval did not go into effect for the first student until the 25th. Therefore, though we used approximately 140 student hours for the project in January, only about 80 were covered by the grant funding.
Amanda, Jeremiah, and Jody met with Jason Battles, Head of Web Services, to sort out issues related to the promotional web display and the usability test. Amanda developed textual content and located a potential image of Cabaniss for the website, and Jody began a section of the Digital Services wiki for Cabaniss information, to be linked into the website.
Amanda selected the Jemison Papers collection as being similar to Cabaniss, for the usability study. Jody analyzed the finding aids for Cabaniss and the Jemison collections and proposed questions for the study.
Jason provided a contact at the IRB and the names of others in the organization who have successfully navigated those waters. Jody contacted these folks for copies of their successful submissions, to use as models for our own. Amanda developed the consent form and preliminary IRB application form. The development of this took longer than we thought. Amanda, Jeremiah, and Jody met to finalize the IRB application on Friday January 29th. We hope the actual submission will be in early February.
- Total files digitized by the end of January, 2010: 6485
- Total files online by the end of January, 2010: 4561
Amanda was out for most of a week this month, so we were challenged for the first time to cover her hours during an absence, but this went pretty smoothly. More critical is that her assignment to this project has heavily impacted the speed with which our other content is getting online, and we've had to alter our workflow and assignments there, incorporating students into the quality control process, which I consider a welcome improvement.
IRB application work has continued to hit new barriers, as we discover more and more forms and hoops to jump through. The latest one is that each person involved in the usability study must pass a 3-hour online certification course on Protecting Human Subjects. We did not plan for this, so that's a good chunk of work hours that we must dedicate to this additional preparation. In addition, one of the students currently working on the Cabaniss project will graduate in May, and thus will not be here when the study takes place. We will train a couple of additional students and use them as necessary when the time comes.
The web page is not yet live, due to Web Services department overload, but Jody developed a press release, and as soon as the website is live, we will send it out to audiences selected by our administration, as well as to appropriate listservs. The preliminary selection includes: Archives and Archivists, Metadata Librarians, LITA, Diglib, and JESSE.
Another anomaly encountered was the discovery that one box of Cabaniss is solely composed of empty envelopes, which had become separated from their letters at some point. We debated the value of digitizing them, but since we agreed to digitize the entire collection, we are forging ahead. Within the box we did discover a couple of letters, which belong elsewhere. We are separating those out and returning them to the archivists for assignment to another box and folder, at which point we will digitize them.
Digitization continues apace. 3801 more images were digitized this month, and 3617 more images went online: the first 14 boxes (out of 77) are done!
- Total files digitized by the end of February, 2010: 10,286
- Total files online by the end of February, 2010: 8178
Early in March, we had some difficulties with the Captureback overhead camera (60 MP digital back with Phase One camera and 80mm lens), which seemed to have taken on a mind of its own (a phone consultation with the company indicates it needs repairs; we cannot afford downtime, so this may have to wait until after the project). This cost us a few hours, and while Amanda and Jeremiah struggled to sort out its issues, Jody analyzed student productivity rates for February, as we did not produce as many images as expected. An analysis of current production rates of all students, not just the ones involved in the grant, revealed that our expectations were a bit high. We had planned on flatbed digitization rates of 17 images per hour, yet for most students, including one of those in the grant, digitization speed was closer to 12 images per hour. Jeremiah and Amanda observed the students' workflow to identify problems and potential need for reeducation or remediation. However, no such issues were noted. It appears that the original rates specified in the grant had been based on experienced students working with the best possible material. The documents under digitization now are more fragile and require more care.
We had a staff meeting to discuss potential solutions and workarounds, and Amanda Presnell as Project Manager made the final selection, as she will be directly responsible for implementation. Students' level of awareness of the speed issue will be raised, and they will be trained to assist in quality control, which will free Amanda to spend more time scanning on the Captureback (on which she produces a consistent 26 images per hour). In addition, the students involved in the grant will be trained on the Captureback as well, so they can fill in there when it is free; also on optimization for the Captureback, again freeing Amanda to scan for more hours. We estimate that substitution of Captureback scanning for flatbed scanning for the slowest student potentially could make up the expected shortfall of images in less than 400 work hours.
We will continue to monitor productivity levels and adapt the workflow as needed to ensure success.
Also this month, we made our first press release File:For Immediate Release.docx to UA Dialog, Tuscaloosa News, and the following listservs: Diglib, Archives & Archivists, Metadata Librarians, Libraries and Information Technology Association (LITA), JESSE, and UALIB.
In addition, Amanda Presnell (Project Manager) proposed a poster session File:Poster Abstract.docx for the Alabama Library Association Annual Conference to be held in April 2010, and was accepted.
We completed the National Institutes of Health (NIH) training on "Protecting Human Research Participants," and made application to the Institutional Review Board (IRB) board for permission to undertake the usability study.
Our website is live! Please see http://www.lib.ua.edu/libraries/hoole/cabaniss for a description of the grant project.
- Total files digitized by the end of March, 2010: 14,944
- Total files online by the end of March, 2010: 12,900
This month, Amanda and Jody responded to two separate requests by the Institutional Review Board for alterations to our proposed usability study application; it looks like approval is drawing near.
After the press release last month, Jody was contacted by Barbara Aikens, Chief of Collections Processing at the Archives of American Art (Smithsonian Institution). Their mass digitization of collections has a similar approach: [Collections Online]. Barbara was kind enough to extend an invitation to tour their shop, and on April 14th, Jody met with Barbara to discuss delivery methodology, work flows and usability studies. Major differences between the Smithsonian's approach and ours include that the series descriptions are separate fields in a database for search support, rather than the finding aid being a single online document, and also that there is no separation between items in their delivery, nor a way to store item-level metadata should some become available in the future. Additionally, our finding aids are searchable alongside our item-level described content (a strength of the Acumen delivery software), whereas the delivery in the Smithsonian interface is separate. Thank you so much, Barbara Aikens!
Additionally, Barbara recommended contacting Jodi Allison-Bunnell, the Program Manager at Northwest Digital Archives, who shared the preliminary report on a fairly broad usability study they did last year. The paper from the study is out for publication review. Despite the rather chilling discovery that neither the item-level described content nor the digitized content available via the finding aid provided sufficient metadata for researchers to "make a decision" -- overall, the researchers seemed to prefer access to the digitized content within the context of the finding aid. We hope to build on the results of this study, and are heartened by the results. Jodi also shared the scripts they used for the study, to help inform our own process. Thank you, Jodi Allison-Bunnell!
Jody posted a query to the EAD listserv announcing the project and asking for information about other usability studies which compare the use of item-level described digital content and digital content made available via the online EAD. Knowledge of others' efforts in this area will help to inform the scope and value of our reports on our own study.
Jeremiah is sorting out how best to juggle students to ensure that we have sufficient student labor over the summer, despite vacations and upcoming graduations.
Amanda uncovered a bug in the linking software, that added items out of order, and Jody repaired it. We are using locally-developed software for this process rather than the bulk linking available with Archivists Toolkit, primarily because we did not want to include links to derivatives within our archival finding aid. The finalized EAD will be uploaded back into Archivists Toolkit when we've finished so the archivists will have access to the completed finding aid for future work.
Amanda also presented a poster session at the Alabama Library Association Annual Convention in Huntsville, Alabama on April 15th. The session was entitled: "Leading the Way Toward Low-Cost Digital Collections: The S.D. Cabaniss Digitization Project" (see Cabaniss_Publicity). This helps us to meet our grant performance objectives, which include: "Publicize the project and its methods through press releases, announcements on appropriate listservs, an article in at least two peer-reviewed publications and presentation of the project during at least two professional conferences." Since this model is intended to help institutions with low funding to digitize their archival content, we considered it extremely important to share this at a state level conference, since few of the less-well-funded institutions are able to send representatives to national conferences. Good work, Amanda!
This month we began digitizing some letterbooks in the collection. These letterbooks require a large amount of attention to detail due to their unusual nature. First of all, the pages are very thin, similar to tissue paper, because the original letter was to be placed underneath the page and traced. So, in order to get a readable scan of one page at a time, acid free paper must be inserted in between the pages so we see only the current page when scanning. Also, these thin pages can become very fragile, and often stick together. Digitizing one of these (some are as large as 700 pages) can be quite time-consuming. Also concerning the letterbooks, we have come across some pages of the books that are very faded, almost to the point of being unreadable. We decided to digitize those pages anyway, because we feel like researchers who were interested could load these images into photo editing software and adjust the image to make the faded writing more legible. However, per our usual digitization policy, we are not digitizing blank pages of the books. We are making notes in the Tracking Files whenever a page is not digitized and why (page blank, torn out, missing, etc.).
We are still working out some issues we are having with the Capture One software and camera. Occasionally the camera will just quit working. After trying the usual troubleshooting techniques (changing batteries, turning off camera, restarting software), if the camera cannot be activated, we will just let it sit for a few hours or overnight. Whatever the issue is will then seem to resolve itself. Amanda dealt with some network connectivity issues this month, but after help from the Office of Library Technology, these issues have now been resolved.
- Total files digitized by the end of April, 2010: 20,267
- Total files online by the end of April, 2010: 18,050
This month was marked by the end of the semester and thus changes in the student availability. Both of the two students already on the grant project will be out for long stretches of the summer, and some of that began this month. We discussed how best to manage the grant staffing over the summer, given that we have a limited number of workstations, and at times will have a full complement of students and staff. Jeremiah worked hard to develop staffing options and scheduling. A third student has now been assigned to the grant project, at least for the summer; her steady input of around 30 hours a week will balance out the loss of the other students for long stretches. At the end of the summer, one of the other students graduates, so we plan to hire again this fall.
After responding to queries and requests from the IRB board, the usability study was approved!! Since the best time to test the students is early in the fall semester before they become too busy, Amanda is planning to begin the marketing in late summer to draw in participants. After consulting with the archivists, it was decided to put a card on a holder on each table in the reading room, a flier on the door, and have plenty of copies of Frequently Asked Questions at the desk for the reference personnel to hand out. Additionally, Amanda will connect with Jason (Web Services) to ensure the laptops and Morae software will be available, will post fliers in the targeted departments, and consult with the secretarial staff to add postings to the departmental listservs. Amanda and Jeremiah will encode the usability test into the Morae software this summer as well.
More information about the usability study can be found here: Cabaniss_Usability_Study
- Total files digitized by the end of May, 2010: 26,585
- Total files online by the end of May, 2010: 24,020
Amanda consulted with Jason Battles, Head of Web Services, about the implementation of the web advertisement for usability study participants. After some discussion, they decided to begin the web ad in early August, just prior to the desired testing time. Since students have the most free time at the beginning of the semester, we decided to target late August or early September for the usability test.
Our long-time graduate student worker (and one of our best!) Kate Matheny is working on the grant this summer, since the two students hired for the grant will be out for long stretches. At the end of the summer session, we will need to replace one of them, as she's graduating. Jeremiah is preparing to advertise; in the meantime, existing student workers will fill in as needed to ensure that we don't fall behind.
Also, Amanda had a week-long vacation this month, so we covered for her work as well. The pace of digitization is still up to par, with over 6900 new items digitized in June alone. Online we have content available now for 39 of the 77 boxes. During August and September, we expect the quantity of scanning will drop back quite a bit, as staffing will be involved in the test process, and also in the evaluation.
This month Jody submitted an article to the Journal of Library Innovation entitled "A Model for Low Cost Online Access to Manuscript Materials" to publicize the grant project; if published, this will be the first of two articles promised to meet the grant agreement. The second will follow the assessment of the usability test, and will incorporate those results.
Publicity this month also includes: Jody linked in the Cabaniss website to her homepage with a description (, updated the professors who contributed letters to support the grant application, and NHPRC posted a Facebook announcement about the project.
- Total files digitized by the end of June, 2010: 33,503
- Total files online by the end of June, 2010: 31,650
This month was fairly quiet, with digitization proceeding apace, and Amanda beginning to encode the usability study into the Morae software. Jody altered Amanda's workflow a bit, so that her makeJpegs script is run from the Linux (destination) server across a mount point against the TIFF files residing on the Windows server. This takes her desktop out of the loop, leveraging the greater processing capabilities of the servers, and freeing up her desktop for other work. Also Jody added a script (moveCabanissContent) which copies the TIFF files across the mount point into a deposit directory for archiving, double checks each copy, and then deletes the copy on the Windows server. This frees up more room on the Windows server for digitization without requiring manual pickups and verifications.
Per the grant reporting requirements, Jody sent in the progress report for the first 6 months. (File:Jan2010 June2010.doc)
- Total files digitized by the end of July, 2010: 38,192
- Total files online by the end of July, 2010: 37,826
As it tends to do, "life happened" to the individuals working on this project during the month of August. We expected to keep one of our current student workers and another who had gone for the summer was to come back and continue work on the project. Instead, both of our current student workers were offered graduate assistant positions in their respective departments that were just too good to turn down. So, we lost both of our student workers within the first week and a half of the month! We began the process of interviewing and hiring new student workers, and they are beginning the first week of September. In addition to this loss of workers, Amanda spent most of the month recruiting and scheduling participants for the Cabaniss Usability Study. This testing began on the last two days of August and will continue into September. Our number for the month are low due to the loss of student workers and Amanda's commitments to the usability study, but thankfully we have previously been very productive and had some cushion to fall back on. This should not delay the completion of the project.
- Total files digitized by the end of August, 2010: 39,633
- Total files online by the end of August, 2010: 39,666
This month both of the new student workers for the collection began work. Each were trained and are now on a regular weekly schedule. They are doing a great job so far. Extra time was dedicated this month to training and supervision as they became acclimated to the collection. While two boxes of files were uploaded this month, none have gone live yet. During the indexing process, a bug was uncovered in Acumen software that is affecting this collection only (perhaps related to the large size of the collection). Our software programmer is looking into this issue currently and hopes to have it corrected in short order. Quite a bit of time was devoted to the Cabaniss Usability Study this month. Recruiting and scheduling of volunteers continued. We ended up with 21 total participants. I (Amanda) did have some issues with people volunteering and then failing to get back in touch with me or scheduling a time and then failing to show up. So after the original ad had been taken down, I ended up doing a second run recruiting a small number of volunteers via departmental list-servs. The study consisted of asking participants to perform a total of 8 tasks, 4 within the Cabaniss collection, and 4 within a similar collection which was laid out in our "traditional" fashion. After the participants performed these 8 tasks, they were asked to fill out a survey which gave them the opportunity to give us feedback on their experience and their opinions about the Acumen software in general and the Cabaniss collection in particular. Initial results from the surveys have been compiled, but there is still more data to be analyzed and compiled.
- Total files digitized by the end of September, 2010: 41,877
- Total files online by the end of September, 2010: 40,755
This month Amanda consulted with Jeremiah to determine how best to use the remaining money/hours on the Cabaniss project. Part of doing that was prepping all the remaining boxes to determine the number of remaining scans. We hiring a new student and adding an existing student to the project. As the project winds down, some of the students will be moved off the grant and onto other supporting funds. I (Amanda) am confident that we will finish the project in a timely manner.
My other major undertaking this month was work on the usability study. I evaluated the results we collected, looking for patterns or other information I could extract from the survey information. I also read several articles or similar studies, in hope to better understand the kinds of things we will want to address in our article.
I did a bit of scanning on the overhead this month, and also trained Dan to use the Capture Back. This will probably mean that he will take over most of the remaining overhead scanning, while I continue analysis on the usability study results. Total scans created this month: 2472.
While prepping the final boxes I came across a large folder labeled “Miscellaneous Documents” that was not on the EAD. I showed it to Donnelly who discovered that most of these documents actually belong in other folders. She is currently working on getting these documents organized, and then we will scan them and place them in the appropriate places on the EAD.
- Total files digitized by the end of October, 2010: 44,349
- Total files online by the end of October, 2010: 43,208
This month we completed Cabaniss scanning! Each of the student workers on the project has transitioned to other funding work. We completed the miscellaneous folder mentioned in the previous month, after working with Donnelly to decide where the items should be located. Amanda spent a large chunk of time this month on QC, including studying the online presentation of the collection for any problems or mistakes. This month we digitized 2219 files.
Jody gathered and analyzed many articles for a background lit review and determined that we will measure learnability by the amount of time to first click plus time to find it and how many correct; effectiveness by the number of correct completions; efficiency by how long to find it and how many steps; and satisfaction by questionnaire responses. The latter will also be sifted for requests/suggestions for change, which will be considered for modifications to the website.
Additionally, Jody is beginning to lay out the specifications for supporting scripts to enable others to emulate this model simply and cheaply without adopting Acumen.
- Total files digitized by the end of November, 2010: 46,663
- Total files online by the end of November, 2010: 46,473
Amanda began to analyze the usability videos in depth, measuring the time to first click, the amount of time to completion of each task, the number of interactions with the computer on each task, and more. She is organizing all the information into spreadsheets along with comments, responses to questions, and background of the participant.
Jody also reviewed several of the videos and began extensive analysis of the data collected during the study. She also continued to review related literature. Additionally, she began do develop software to enable others to use this model outside of Acumen, developing specifications and configuration files as well as a file-testing script to ensure that defined specifications are met. Jason was asked to develop a static HTML template page to be used for generating access pages for both items and folder content.
- Total files online by the end of December, 2010: 46,663
DIGITIZATION COMPLETED AND ONLINE!!
January and February 2011
During these two months, we finalized analysis of the usability study and developed an article with the help of Kevin Walker, Assessment Specialist for UA Libraries. This paper was submitted to JCDL 2011 (Joint Council on Digital Libraries annual meeting), in hopes of presentation there this summer. Additionally, with the help of Will Jones (under the direction of Jason Battles, Web Services), we developed software to support implementation of the use of this model of digitization regardless of the finding aid delivery method. Thus, use of this model is not software-dependent. Institutions currently hosting finding aids via other software packages can still benefit from this method of digitizing manuscript content. More about this can be found here: Software_For_You!.
Spreadsheets with digitization progress, updated monthly:
Numbers digitized each month and number online, with running totals: File:TrackByMonth.xls