Dissertation | Literature review < Proposed technique > Commentary

Auditing catalogue quality by random sampling

3. The proposed technique

There are two stages to the audit: the catalogue-to-collection test which evaluates the accuracy of records and the collection-to-catalogue test which checks for uncatalogued items and, incidentally, duplicate records. The tests are designed to be performed at the same time.

First, a random sample of catalogue records must be produced by the computerised library system. What constitutes a random sample and alternatives when one is not possible are discussed in the next chapter. To compute the number of records to include in the sample, it is necessary to decide on the acceptable margin of error in the result and to make a rough estimate of the proportion of errors that will be found.

The worksheet (3.1) has been designed with books in mind but should be adaptable for different requirements, bearing in mind that gathering more detail on errors takes more time. Each type of error is assigned a code so that a single copy of the worksheet can be used to annotate printouts of full records. Alternatively, the worksheet can be copied as many times as there are records in the sample and staff can tick the relevant boxes.

For each record in the sample, find the corresponding item in the library and also the item five places to its left. Using the worksheet, check the cataloguing of the first item. If it cannot be found and is not on loan, note its classmark and search for it on a second occasion; if it is on loan then recall it for the cataloguing department. Methods for dealing with items that persist in their absence are given in the next chapter. Next, search on the catalogue for the second item, by control number and then by other keys. Make a note if no record can be found or if more than one record for it is found in the process of searching.

It may be more efficient to locate several items at a time and send them elsewhere to be checked rather than considering them individually, especially if staff locating the items are not trained in cataloguing, but this depends to some extent on the physical arrangement of the library.

The results can be displayed in the table given on the worksheet, or entered in a spreadsheet or database for further manipulation. Simple measures are the proportion of records with errors (and the proportion with more than one error), the number of uncatalogued items in the collection and estimates of the number of duplicate records and missing items. One straightforward division of errors is to consider those in access points as having more impact on retrieval.

3.1 Worksheet

Catalogue-to-collection test

Control number: ______________________________________

If the catalogue record lists no copies held for this item, record code Z or tick this box [ ]
and go to the collection-to-catalogue test for the item five places to the left of its classmark.

Please tick a box for each field in the catalogue record, or note the codes against each record (for example T2 M4 R1). For author and subject headings, consider only whether they match the forms given in the authority file.

acceptable incorrect or incomplete missing not applicable code
Title         T
Material description         M
Statement of resp.         R
Author heading(s)         H
Edition         E
Physical description         P
Imprint         I
Series         S
Classmark         C
Subject heading(s)         D
Language         L
Genre / category         G
Location / branch ID         B
  1 2 3 4  

If there are other errors in the record (for example absence of uniform title) then list them below but do NOT count them in the statistics.

Collection-to-catalogue test

Control number: ______________________________________

present (P) missing (M) duplicated (D)

Dissertation | Literature review < Proposed technique > Commentary

Owen Massey McKnight <owen.mcknight@worc.ox.ac.uk>