Separate but equal? A system comparison study of MEDLINE's controlled vocabulary MeSH

This study tested the effect of controlled vocabulary search feature implementation on two online systems. Specifically, the study examined retrieval rates using four unique controlled vocabulary search features (explode, major descriptor, descriptor, subheadings). Each search feature was applied to nine search queries obtained from a medical reference librarian. The same queries were searched in the complete MEDLINE file on the Dialog and Ovid systems. The unique records, i.e., those records retrieved in only one of the two systems, were identified and analyzed. Dialog produced equal or more records than Ovid in nearly 20% of the queries. The study demonstrated that users need to be aware of system-specific designs that may require differing input strategies across different systems for the same unique controlled vocabulary search features. The paper concludes by making recommendations and suggestions for future research.