Finding our foundation: Analysis of the LISA database for research retrievability

Objective: The primary objective of this study is to examine the Library and Information Science Abstracts (LISA) database to determine if research literature can consistently be retrieved by using keywords identifying the research methodologies used.

Question: For the journals named, are articles identified as ‘research’ able to be consistently retrieved by using keywords related to research methodology? Methods: Citations from the top 10 Library and Information Science journals for 2001 as identified by Koufogiannakis, Slater & Crumley (2004) were obtained, then a filter developed by Catherine Beverley (2004) was used independently to identify research articles. The resulting sets of citations were compared, then the two datasets were analyzed in order to consider retrievability and fit for assigned keywords from the LISA database.

Results: Although it would need to be tested against a random set of citations rather than the purposive sample tested here, our analysis suggests that retrieval using the descriptor terms alone may succeed in only 31.5% ± 5.2% of attempts, with a 95% confidence interval.

Conclusions: The LISA thesaurus is not consistent or sufficiently comprehensive to serve the needs of researchers. Recommendations for the improved retrievability of LIS research literature from the database are made.