Retrospective Application of Subject Headings, Part 2-A Case Study at the Central Washington University Library

Article excerpt

ABSTRACT

This article describes implementation of a methodology followed for the retrospective application of 474 subject headings to already existing bibliographic records. Findings and considerations for future application are discussed.

INTRODUCTION

The purpose of this article is to describe the results and some of the ramifications of a project to apply subject terminology retrospectively. As outlined in part one of this article (1), a list of subject authority records (SARs) from 1995 in the Central Washington University Library were extracted and printed out. The headings and references were then searched in the online catalog, CATTRAX, and suitable records were updated with the terminology, subdivided as appropriate. The number of headings reviewed in this study was 474. No selection criteria were used to eliminate headings not likely to produce results.

The development of the procedure used was discussed in the first part of this article. The results to be discussed are from following that procedure as it is described here. The number of changes made to bibliographic records as a result of this process have been significant (greater than 1000). Local changes in subject authority records (SARs) have been noted as well, many of which will be submitted as proposals via the Subject Authorities Cataloging Project (SACO) through the Library of Congress.

Procedure:

Step 1. Searched each subject as a phrase in technical mode and then recorded the number of hits

Step 2. Searched each heading string as a word search (in general, omitting terms used in qualifiers)

Limited by title words of the subject string

Browsed for appropriate titles

Applied the new subject headings if appropriate

Step 3. Searched at each see from reference as a word search

Truncated plural forms to eliminate re-keying and to increase recall (or searched by plural and singular form when recall was too large). I limited results by words in title of the new heading or the reference

Browsed for appropriate titles

Applied the new subject headings if appropriate

Step 4. Searched each broader term from the SAR Limited results by words of the new heading in the title If the broader term were subdivided, searched it also in its un-subdivided form Searched ad-hoc broader terms, if they were used

Browsed for appropriate titles

Applied the new subject headings if appropriate

Step 5. If alternative terminology were appropriate, did a word search on that terminology

Limited results by words of the new heading in the title

Browsed for appropriate titles

Applied the new subject headings if appropriate

Step 6. Searched each heading again as a subject string

Recorded the number of hits, from which the original number of headings was deducted. The net results are recorded as the primary changes made to the bibliographic records.

Limiting by word in title was done only on the occasions when recall was too great to make browsing the file effective. In most cases, retrieving fewer than fifty or so records still allowed browsing, so limiting was not used.

During this process, I attempted to resolve appropriateness of the new heading for the work by searching other catalogs or having the item paged from the stacks, when either step was needed. This was rarely done.

Step 7. Additions of additional subject headings were tracked as secondary changes, e.g., corrected headings, additional headings for other related topics, etc. Stylistic changes or those unrelated to this specific process were not counted. For example, when examining the heading for the fictional place, Thalia, Tex., upgrades to other records for the works of fiction by Larry McMurtry were not counted in this study.

Step 8. Potential SACO applications were also identified. …