Observing users to validate models

User models serve two purposes: to help us understand users, and hence determine how to supply them with effective search services; and as a framework against which to evaluate the quality of those services once they have been developed. In this extended abstract we describe an experiment we have undertaken in which we observe user behaviours, and try to determine whether these behaviours can be connected to search quality metrics via an existing or novel user model. We provide summary evidence that suggests that the answer is a qualified "yes".