Content Evaluation of Textual CD‐ROM and Web Databases

Citation

Parry, F. (2002), "Content Evaluation of Textual CD‐ROM and Web Databases", Online Information Review, Vol. 26 No. 4, pp. 278-278. https://doi.org/10.1108/oir.2002.26.4.278.1

Information professionals today face an ever‐growing number of databases and subsequent calls on their budgets, with publishers sometimes making extravagant claims to entice libraries to buy or subscribe. It is all the more important, therefore, that these databases are carefully and systematically scrutinised, in order to ensure quality, relevance and value for money.

This book takes us through the database content evaluation process, with chapters on subject scope, database dimensions, source coverage, record content, accuracy and reliability, completeness, quality of indexing and abstracts, cost. In each of these the author proves himself to be a diligent and meticulous researcher. He has a keen eye for detail, which is manifested in numerous well‐researched database tests. Jacsó concentrates mainly on a baker’s dozen of databases well known to librarians and information professionals, although the principles he employs are also applicable to other databases. Key among these are Library and Information Science Abstracts and Information Science Abstracts, representing the bibliographic databases; and Ulrich’s, Gale’s and Bowker’s Complete Video Directory representing the directories.

Jacsó’s evaluations are scrupulous and fair and are accompanied by screenshots to illustrate his points. Although he highlights both positive and negative elements of database content, Jacsó is at his best when he points out the absurdities and idiosyncrasies of some databases. One such example can be found in the chapter on accuracy, where the Mental Health Abstracts database is reported as having managed to find 25 different ways of spelling the word English in its language index. Most databases are prone to inaccuracies and typographical errors, and in such cases many records are irretrievably lost or hidden. Inconsistent, incomplete and poor indexing and abstracting services in a database blight that database’s effectiveness.

The author also performs a valuable service by alerting us to the ways in which a database can appear to be bigger than it really is, more “international” or broader in its coverage. Unlike most database errors, these are instances where an alert searcher can thoroughly test the claims of a database publisher. Examples include the different methods databases employ when recording journal title changes: successive entry or latest entry cataloguing; partial versus full indexing of certain titles; inadequate coverage caused by late take‐up or early termination of a journal title. Jacsó is quite forthright in his opinions and takes care to back them up. An example of this can be found in his thorough dissection of the database source coverage claims made by Information Science Abstracts. Here Jacsó finds a “disharmony between the claimed scope and actual coverage of the ISA database”. The image that most readily comes to mind in instances like this is a dog with a bone!

Jacsó has succeeded in producing a fine methodology for evaluating database content. He says in the introduction that he has not considered hardware and software issues, which is understandable, but also a pity since I am sure he would have had many useful insights into database interfaces. Another book perhaps?