Bibliographic Management

BNB Currency Survey

This survey, measuring the currency (availability at the time required) of BNBMARC records of post 1974 published titles with UK imprints currently being acquired by UK libraries, has been carried out continuously at the cataloguing stage since January 1980, and at the ordering stage since January 1988, until it was ended on 31st March 2005.

Every month a rolling, random selection of 6 academic and 6 public libraries each provided a random sample of 10 titles they were about to catalogue and 10 titles they were about to order. The samples were returned to UKOLN, where they were checked against BNB files on the British Library database. Data from the search results were input into an Access database to enable statistical analysis to be carried out. This analysis produced the index of currency or 'hit rate'. Results were conflated over a 12-month period to avoid fluctuations caused by random and seasonal variations. A second search was carried out 12 months later, on titles for which records were not found on first search. The data from the recheck produced a revised hit rate.

The results of the survey demonstrate that, in the last few years, the British Library has significantly improved the currency of records on its database at the time they are needed by libraries.

The study also recorded data on other elements which may have affected the hit-rate (eg. CIP records, CIP record purchase, British Library cataloguing initiatives including the Copyright Libraries Shared Cataloguing Programme).

Cataloguing stage

At the cataloguing stage, hit-rates of 63-65% were recorded between 1981 and 1987. Thereafter the hit-rate rose steadily to 89 % in July 1994, declined to 81% by November 1995 after which it rose again. The hit-rate was 90% in January 2005, with a recheck hit-rate is currentlyof 92%.

Hit rate at date of sample

Hit rate six months after sample date

Ordering stage

The ordering stage survey produced an initial hit-rate of 74% in 1988. It then rose to 84% in May 1995, then declined to 79% by January 1996 after which it rose again. The hit-rate was 87% in January 2005, with a recheck hit-rate of 91%.

Hit rate at date of sample

Hit rate six months after sample date

End of survey

In March 2005, the decision was taken to end the survey. Although the methodology remained valid, a review of the survey revealed the following:

• Until recently only a few libraries declined to take part. However, in the last few years it has taken more time to find libraries able to take part and there have been gaps in the sampling schedule. Reasons given for non-participation are:
  • Installation of new or upgraded library management system.
  • Staff restructuring.
  • Staff shortages and illness.
  • Cataloguing and / or acquisitions out-sourced, and the supplier not able to provide samples.
  • Distributed acquisitions and cataloguing at multi-site academic libraries making it difficult to provide samples that accurately reflect the total range of subject areas of an institution.
  • And occasionally, running out of bookfund part-way through a financial year.
• In recent years there has also been an increase in the number of samples which either did not contain the full number of titles required or came back marked as 'nil return'. This is because:
  • There may not have been enough material going through acquisitions and cataloguing on the sample date.
  • On some occasions libraries reported that they had used all their bookfund before the end of the financial year. This reflects the disturbing trend of bookfund cuts.
  • Staff shortages occurring after a library has agreed to take part, due to illness and vacancies. Often there is no-one else who can take on the survey returns in the interim.
• Occasionally titles needed to be removed from the samples if they were identified as:
  • Titles published pre-1974.
  • Titles published outside the UK and without a UK distributor. It appears that increasingly publishers list all their publishing locations on the title page, but may not actually publish an item at all of the locations.
  • Although the size of the sample means that occasional gaps are not a problem, this started to happen frequently enough to raise concerns over the statistical validity of the survey.
  • Until 1990, samples were near the maximum of 120 titles per month, whereas in 2004 the samples ranged from 62 to 89 each month.
  • Although hit-rates are calculated over a year, at some point the lower numbers and the accidental inclusion of non-UK publications will mean that there is a risk of biased samples.
  • However the methodology remains valid, as long as sufficient samples of accurate data can be collected.

Apart from all the above difficulties, there now appears to be little interest in the survey. Analysis of the UKOLN website pages where the results are published shows few, if any, visits to these pages.

Other information

UKOLN staff who worked on the surveys: Ann Chapman, Alison Holbrook, Janet Kinsella and Alan Seal.

An archive version of the BNB Currency Survey site is available on the Internet Archive. On 13 March 2013 http://www.ukoln.ac.uk/bib-man/surveys/bnbmarc/ had been crawled 24 times going back to 31 January 2003.

Publications

Chapman, A., 1999. Availability of bibliographic records for the UK imprint. Journal of Documentation, 55 (1), pp. 6-15. Opus ID: 23779

Chapman, A., 1997. Bibliographic record provision in the UK. Library Management, 18 (3), pp. 112-123. Opus ID: 23786

Chapman, A., 1995. National library bibliographic record availability: a long-term survey. Library Resources and Technical Services (LRTS), 39 (4), pp. 345-357. Opus ID: 23787

---

Content by: Ann Chapman of UKOLN.
Page last revised on: 25-Mar-2013
Email comments to: web-support@ukoln.ac.uk