Collection surveys examine the condition of collection pieces, display environment, and storage quality. When assessing for damage and possible conservation treatment, a quantitative survey will supply the administration with facts to evaluate the loss of knowledge and secure financial commitment for a course of action. For large collections, a survey can be time and resource intensive. A simple easy to read form should be created for the survey to help focus the information being gathered. Ideally, the form will be one no longer than one page and designed for easy manual recording. Descriptions for each category should be precise in order to avoid misinterpretation. Adequate space should be left for recording miscellaneous information that has not been accounted for. Once the form is completed, a test run should be done a random sampling of the collection to ensure that the form will work in the actual survey. This allows time to revise the survey before implementing the study. "There are six distinct phases in a survey, and sufficient time needs to be allotted to each of them: agree statement of survey purpose; record the purpose of the survey; describe the collection and define terms; undertake pilot survey; collect data and survey itself; analyze and present the data."v
The pilot survey should have a pre-determined time limit. This will provide the information necessary concerning time allotment for the actual survey. Information that can be derived from the pilot include number of collections or record groups surveyed, number of individual objects surveyed, total number of collections or record groups, approximate number of objects in each collection or record group, and maximum number of objects that can be surveyed within time allotted.
The survey design should be tailored to collect specific information while reducing unnecessary information. The format should focus on gathering numerical not descriptive data. Taking some time to get familiar with collection materials will help identify elements to track such as number of bound manuscripts, number of document boxes, number of bound objects, etc. Staff training will involve relaying purpose of the survey, the history and nature of the collection, general preservation and conservation concepts and relevant terminology.
Large collections will rarely have the staff and resources to be fully surveyed and will require a sample survey for the entire collection. "The number of items that will actually be examined is determined by a statistical formula that is based on the size of the collection and on the level of confidence and accuracy desired. Every collection has it's own unique characteristics. The wisest course would be to discuss the sampling scheme with a statistician, so time and money are not lost through oversight or error."xi