Submitted by Mike McGraw, Cleveland Health Sciences Library,
Case Western Reserve University, Cleveland, OH
My poster was billed in advance under the title "Two Methods for Limiting PubMed Searches to Nurse-Authored Research Articles" but appeared at the conference with the title "Can I Beat PubMed's Limits at their Own Game?"
It reported on a kind of case-control study I did where I compared the effectiveness of two methods - one using PubMed's "Limits" tab, the other a Boolean clause of my design - at filtering subject searches down to articles that were both nurse-authored and research-based. My desire to improve on existing methods for doing this was based on repeated requests for such articles from nursing students at the institution where I am a reference and user services librarian.
Three subject search "stems" were linked via a Boolean "AND" to each of the two filters described above, and PubMed was queried based on these six search sentences. I pulled all of the resulting articles and kept them in their six sets.
Having decided on predetermined definitions of both "nurse-authored" and "research-based," I determined what proportion of each of the six sets of articles qualified on both scores. A comparison of each "case" (my Boolean filter) to the "control" (PubMed's Limits) for each subject search showed that my "case" method produced a higher proportion of qualifying articles all three times. The differences were unlikely to be due to chance alone based on a simple, well-known statistical test.
Details on the search queries used can be found in my article, McGraw M. Narrowing PubMed searches to nursing-related articles. Comput Inform Nurs 2009;27(5):272-5. [PMID: 19726919]. The statistical results were in the poster but are not in the CIN article. If you'd like more information, please email me at michael.mcgraw at case(dot)edu.