Authors: Elisabeth Aurelle, Michael Conkin, and Tonette Mendoza
Note: This is part 2 of a 4-part document. Move among documents using links at the end of the document being viewed.
III. Cataloging Functions
A . Searching for Bibliographic Copy
The first step in most cataloging is searching OCLC and/or RLIN for a usable bibliographic record describing the item in hand. In some cases, cataloging with copy is performed at the time of searching; standards for these two functions cannot always be c ompletely separated.
Perform the fewest searches needed to achieve definitive search results, and no more. In practice, a minimum of 3 searches (if no hits are found in the 1st and/or 2nd search) per item suffices. In any case, 5 searches are probably enough. Experienced catalogers use judgment in searching for copy.
As of this writing, no methods for evaluating the performance of this function are in place. Performance indicators for searching for copy as a distinct function might include:
B. Cataloging with Copy
This function usually begins after successful searching has identified a record (or records) with usable copy describing the item in hand. Catalogers using copy add, delete and/or modify data to reflect local cataloging and holdings practices to produce m achine-readable records for addition to the GLADIS database. Aspects of the performance criteria for this function are related to those described in searching for copy.
The single most effective performance indicator to measure correctness/error rates in these categories is a regular post-cataloging quality review check comparing the GLADIS record with the item cataloged. As of this writing, no methods for review in these categories are in place.
RECOMMENDATION: The panel recommends that material cataloged via the utilities not be released for shelving until after the record appears in GLADIS
Additional indicators might include:
In practice these are not truly measurable indicators, since no overall base figures and no figures for reports/requests not sent will be available for comparison; if tallied and categorized, perhaps year-to-year comparisons may yield quantifiable data . These reports may legitimately be used as anecdotal pointers when evaluating a unit's performance, however.
RECOMMENDATION: The panel recommends that an improved method (possibly web-based) be developed for end-users, especially primary clientele, to report catalog problems, and an improved response method for getting back to them directly (with, e. g., a FAQ section)
Additional standards for particular categories of cataloging using copy and more specific standards for evaluation of quantity and timeliness of this function may be developed locally.
C. Original Cataloging
Original cataloging results in the creation of a bibliographic record directly online in GLADIS or via a cataloging utility, and is performed when usable copy for the specific item in hand is not available at the time of cataloging. In general, original c atalogers:
Within this broad context, original catalogers are allowed maximum flexibility and latitude in applying rules and rule interpretations to specific cases, and are expected to subject their work to exacting self-review.
This function covers the range from "pattern" cataloging (which uses elements of already-cataloged related-item records such as different editions, earlier titles, supplements, etc., in the creation of a new record) to permanent brief cataloging (which restricts cataloging to minimal bibliographic elements) to full original cataloging.
NOTE: In addition, the central unit participates in LC's BIBCO (formerly NCCP) program, contributing full-level bibliographic records for UC-owned monographs to the LC database.
Original cataloging typically requires:
NOTE: Authorities work is generally considered to be a component of original cataloging; these standards treat that function separately (see III.D., below)
The most effective performance indicator for evaluating original cataloging would be a regular post-cataloging quality review check comparing the GLADIS record with the item cataloged, concentrating both on the mechanical aspects of cataloging as well as the intellectual aspects, and taking fully into account subjective considerations and judgments made at the time of cataloging.
In practice, such reviews may best be undertaken as part of the rigorous self-review characteristically exercised by catalogers performing at this level.
|RECOMMENDATION: The panel recommends that material cataloged via the utilities not be released for shelving until after the record appears in GLADIS.|
Additional indicators might include:
In practice these are not truly measurable indicators. These reports may legitimately be used as anecdotal pointers when evaluating a unit's performance, however.
|RECOMMENDATION: The panel recommends that an improved method (possibly web-based) be developed for end-users, especially primary clientele, to report catalog problems, and an improved response method for getting back to them directly (with, e. g., a FAQ section)|
|RECOMMENDATION: The panel recommends that all normal-priority material likely to be held in an original cataloging backlog be represented in GLADIS by an in-process record; if that is not practical, at the very least those items likely to be he ld in such a backlog for 30 days or more should have an in-process record in GLADIS (this does not include material set aside by policy or category).|