One information retrieval tool used by information professionals is the methodological search filter to identify specific study designs such as randomized controlled trials. Search filters can help to optimize retrieval by maximizing sensitivity (i.e. identifying as high a proportion as possible of relevant records) whilst achieving adequate precision (minimizing the number of irrelevant records to be assessed). As search filters proliferate, it is unclear how searchers evaluate and choose between available filters.
A series of reviews was undertaken to establish the methods that have been used to measure and report search filter performance. A web-based questionnaire elicited information professionals’ views on what measures are most useful in choosing between available filters.
Filter performance comparison studies most commonly reported highest sensitivity, highest precision and optimal/balanced filter strategies. Tables were the most frequently used method of reporting the results of filter performance comparisons but graphs may be more useful. Respondents used a variety of ways to help in selection: not only referring to reported performance (sensitivity and specificity in particular) but also relied on advice of colleagues and the providence of the filters.
Results from this study will inform the development of a search filter performance website.
Cynthia Fraser; firstname.lastname@example.org