Show me your search strings! Statistics and user behaviour

Abstract

What can statistics tell us about user behaviour in libraries? This presentation looks at search strings in particular and analyses search behavior with the purpose of improving library services. We have studied statistics from our federated search engine, and will focus on the most common mistakes that these statistics tell us users make. Examples include using too wide search terms or searching for topics that are not covered by a particular database, for instance searching for chemistry related subjects in a humanities database. The purpose of studying the mistakes searchers make is to give us as library workers a better understanding of how our users think when they go through a search process, and thus making us better at adjusting our information literacy classes for their needs. This knowledge can also play a role in the way we present our discovery services to our users, and it may influence our purchase policy and perhaps enable us to provide publishers with feedback on what our users find difficult when searching. This project also tells us a something about the discrepancy between some users’ perceived search competencies and what they actually do when they enter a database. We find that by looking at more than the number of searches and downloads, we learn valuable information about our users, and believe that using statistics in this way will help to improve library services.