Yale University Library Collection Development Council

PROPOSAL FOR REPLACING MISSING BOOKS
9/2000

Backgound: The working group (Will Wheeler, Sue Roberts, Jeffry Larson, Sue Crockford-Peters, Joan Swanekamp, Audrey Novak ex officio advisor) met numerous times between March and August and also consulted with Systems staff.

In late September 1999, the Systems Office produced two extracts from Orbis of missing items (1395 items SML MSNG) and not-returned items (1962 items SML NTRN). The Systems Offfice lacks the capacity to do additional programming on these files to sort them by selector and recommended that the files be handled in Access. Members of the group with Systems staff planned a set of queries that could be run against these files to sort them according to selectors' responsibilities. The parameters of the queries were laid down by Jeffry Larson and Step Schmitt. The files were left as unfinished business of CDC; when the issue was reintroduced, Sue Roberts and Will Wheeler volunteered to explore the Access alternative.

A new extract was generated in July 2000 based on data 18 months back and older. THese files were SML MSNG (2439 items) and SML NTRN (3566 items). Initially the group decided to do a trial using the queries that had been designed to test both the queries and the processes that would need to surround the replacement process. Jeffry, Sue, and Will agreed to be the testers on titles in their areas. We hoped that Ann would be able to provide funds for a student expert in Access to write the queries and help run them against the files.

Ann made a counter-proposal: that we treat replacements in a more routine and automatic fashion without the necessity of further programming, complex querying or selector review. Using central funds that Ann would identify, we might thus draw down the backlog of missing/lost books over a period of perhaps two years by automatically ordering replacements. We could establish a relationship with an OP vendor such as Alibris to supply titles automatically within specified price limits and to report on higher prices and unavailability. Selectors would deal with problems but not every title.

The committee considered the alternatives and discussed pros and cons. Following is a proposal in two parts preceeded by a brief description of the extract.

An initial breakdown of the SML files (2439 MSNG and 3566 NTRN = 6005 items) shows that 1494 of these are level 9 record with no bibliographic records. We cannot process these until the titles have been reconned . Thus, the real usable total is (6005 - 1494) = 4511.

Of the 4511 usable records, 3342 (74%) are in English, and 2148 (47%) are in English and published after 1979.

THE PROPOSAL

1. Automatic replacement. A pilot project would attempt to replace either the 3342 English language titles or the 2148 English language titles published after 1979 using the services of a vendor. The committee is more comfortable with these parameters because they are probably higher use titles, are more recent (if we use post 1979 criterion), more readily replaceable given their age and language , and would not need further decision-making about duplication (this decision having already been made once).

Decisions will still need to be made about the extent of involvement of SML's Order support team. Do we create OPRs right away (a costly process) or do we simply give the file to a vendor directly, simply creating a copy statement for the potential replacement? Negotiations will need to take place with vendors to determine if they have the skill sets to replace exactly the missing titles, and to set the terms of a relationship.

Processes will still need to be developed:

Titles will need to be searched by Access Services.
Record and catalog maintenance will need to be done by Access Services and the Catalog Maintenance Team. Staff will suppress the missing copy, but Orbis records will indicate to the public that a replacement is on order.
Receipt procedures will need to include bibliographic discrimination and maintenance of statistics. These will require training perhaps one or two D-level staff.

2. Strategies for more complex replacement decisions. We propose to continue working on a long-term strategy by instituting a pilot using Access queries as envisioned above to identify non-English language titles in Will's areas, a process that would eventually scale to all selectors. Once titles in their areas of responsibility are delivered to them, selectors could determine which items might best be withdrawn and which should be replaced. Some selectors may choose to use a more routinized method similar to what is proposed above, where Acquisitions or a vendor could take a title list and try to find replacements (for example, German language items might be sent to Harrassowitz, French to Touzot, etc.).

Will will likely experiment with both alternatives: exploring the possibility that foreign vendors can search for OP titles in their language areas, and using Access parameters such as date range (1980-pres.) to see how complex queries really are. For example, may choose to create filters/queries that pull out all gaps where item is a volume in a series or serial or to replace all that have more than x number of circulations, etc.

Respectfully submitted Sept 28, 2000 by Sue Roberts & Will Wheeler

Send comments to ann.okerson@yale.edu