- Preservation Home
- About
- Collections Care
- Conservation
- Digital Preservation
- Emergency Management
- En Español
- FAQ
- Preservation Science
- Resources
- Outreach & Training Opportunities
- Have a preservation question?
Ask-a-Librarian
Related Links
Methodology
Sample Size
Accuracy of survey results generally increases as sample size increases. Sample sizes chosen for surveys may range from less than 1% for large collections (e.g., several million items) to 100%, (i.e., each item in a collection is surveyed). Although a 100% survey is ideal in terms of representing a collection accurately, such a high percentage is often not practical due to time and workforce constraints. For the Pre-1801 book survey, survey staff chose an intermediate sample size of 10% for four of the five collections. Survey staff chose a slightly higher sampling rate of 15% for the Canon collection, because that collection is smaller than the other collections in terms of number of shelves.
Book Selection Procedure
Surveyors counted the number of shelves in each range before the sampling process began. Staff generated random numbers corresponding to shelf numbers, allowing a random sampling of shelves in each collection. Surveyors selected and surveyed every Pre-1801 book on each sampled shelf. This procedure made for a very efficient use of time while in the stacks, allowing a greater number of books to be surveyed in a given length of time.
Survey Form
Survey staff developed a survey form prior to the start of sampling for recording the book size and condition data. In developing the form, staff sought a balance between gathering data of sufficient detail to allow a thorough collection characterization and meaningful conclusions, while avoiding data considered less critical.
Survey staff recorded data on a hardcopy form during the survey, which they later transferred to an Excel spreadsheet containing the same column headings as the survey form. Surveyors then used the spreadsheet to calculate the numbers of different sized boxes necessary to house the books. The spreadsheet also allowed staff to rapidly calculate statistics such as the number of full-style books, the number of books requiring individual housing, and the number of books with damaged outer joints. Staff could also employ the database functions of Excel to generate statistics such as the number of full-style books covered in leather, if necessary.
For each Pre-1801 book surveyed, staff recorded the following data:
- Collection
- Range and shelf number
- Publication date
- Dimensions
- Style
- Cover material
- Condition of cover material
- Evidence of rebinding
- Evidence of previous treatment
- Housing needs
- Condition of boards
- Condition of caps
- Condition of spine
- Condition of outer joints
- Condition of inner joints
- Condition of text block
Based on an evaluation of the above characteristics for each book, the surveyors determined whether the book was in more urgent need of care than most books. Major damage (e.g., partially detached or detached boards, split spine, or loose sections of pages) and/or vulnerability to future damage (e.g., unsupported pamphlet) or theft (i.e., miniature book) were primary factors in identifying books in most need of care. All surveyed books judged in need of more urgent care than other books were relocated to a more secure section of the Law Library stacks. Surveyors flagged all books within this group exhibiting major damage (as opposed to miniature books) on the survey form for inclusion in survey statistics. Surveyors consider flagging books with major damage a very important aspect of this survey because, when planning future treatments, these books will be given priority in a treatment schedule. Treatment staff can use the survey data to estimate necessary labor and materials.
The following photographs show examples of characteristics observed and recorded during the survey.
Vellum spine cover is missing. Expand image |
Vellum binding showing a previous treatment (i.e., a fill and joint repair at the head of the spine). Expand image |
Book with a partially detached board. Expand image |
Full vellum binding with a delaminated corner and warped boards. Expand image |
Damaged cap on full paper binding. Expand image |
Split spine evident along gutter of text block. Expand image |
Torn outer joint at tail of vellum binding. Expand image |
Torn inner joint at tail of vellum binding. Expand image |