Proposed Methodology for U.S. National Libraries RDA Test
March 13, 2009
Purpose of the RDA testing
To determine whether RDA Online in its initial release offers sufficient benefit, compared to current methods, to achieve the objectives that the Joint Steering Committee for Development of RDA has stated for the proposed new cataloging code (http://www.collectionscanada.gc.ca/jsc/docs/5rda-objectivesrev2.pdf [PDF, 59KB]). Based on the analysis of the test results, and considering the cost of implementation, the three US national libraries will determine whether to implement the initial release of RDA.
Assumptions of the test
- Test results will be shared with the community.
- Only RDA Online in its initial release will be tested.
- RDA Online will be tested in existing library systems.
- All data produced in the test will be freely available to members of the US information community for use or re-use.
- In addition to formal testers, others will be invited to use the US National Libraries RDA Test methodology and to share their test results with the LC/NAL/NLM RDA Test Steering Committee and the community.
- System developers will be invited to explore ways in which RDA records behave in existing library systems and/or how RDA might impact development of future systems.
- The analysis of the test results will consider both qualitative and quantitative reports from testers about their experiences in using RDA, compared to the codes they currently apply.
- Approximately 20 test partners will produce records. Test partners will include institutions that produce MARC 21 and/or non-MARC (e.g. MODS, Dublin Core) records; library professional associations; library educators; members of the vendor/book jobber community; and selected individuals. The institutions will include, to the extent possible, small, medium, and large public, research, government, school, and academic libraries and will provide a cross-section of seasoned and novice practitioners at various staffing levels. In general, each participating test partner institution will name at least two individual staff members as testers. All the test partners will report their results directly to the US National Libraries RDA Test Steering Committee.
- Testing will last six months. The test will begin with a three-month training and practice period after the initial release of RDA Online becomes available, currently projected for July 2009 or later. Actual test records will be produced in the second three-month period.
- The Steering Committee will identify a core test set of 20 to 25 resources that together are representative of the range of digital and analog resources acquired, cataloged, and maintained in contemporary US libraries. The core test set will include textual monographs, AV materials, serials, and integrating resources.
- Each test partner will produce an RDA bibliographic record and a bibliographic record produced using its currently adopted cataloging code, e.g. AACR2, DACS, VRA, for each resource in the core test set. No staff member should catalog a given resource using both codes.
- test partner institutions will assign different staff members to produce the RDA record and the record using the cataloging code that the institution currently applies.
- "Singleton," non-institutional test partners will be paired with other individual test partners by the Steering Committee. For any resource in the core test set, each member of the pair will produce a bibliographic record using either RDA or his/her current cataloging code, but not both.
- Test partners will follow their normal workflows to perform the associated authority work in support of both the RDA test records and the test records produced using their current cataloging codes. If an institution does not normally perform authority work, it does not need to supply authority records in the test.
- In addition to the core test set, test partners will also produce RDA records for resources in their usual categories of cataloging, including any areas of specialization, e.g. foreign languages, cartographic material, or music scores. The additional test records will be created using only RDA. The Steering Committee has been advised that a set of 800 records for the entire test will ensure that the results of the test will be statistically valid; therefore each test partner will, ideally, produce at least 25 additional test records.
- For each record, each tester (individual test partners and staff members of test partner institutions) will also complete an online questionnaire that elicits information on the amount of time required to produce the records and on qualitative aspects such as how easy it was to understand the RDA instructions and the instructions in the currently adopted code; what changes in workflow were necessitated by the use of RDA; what stumbling blocks were encountered in using RDA in local systems; and what kinds of consultation with colleagues or other documentation were needed to produce the records.
- As one of the qualitative aspects, test partners will solicit feedback from their internal end users about the RDA records they create.
- The test partners will make available the records they create in this test to the Steering Committee for review and analysis.
For general contacts or for information about the test, please email:
Susan Morris, Special Assistant to the Director for Acquisitions and Bibliographic Access ([email protected])