The Hawthorne studies in the 1930s demonstrated how difficult it is to understand workplace behaviour, and this includes professional performance. Studies of interventions to improve professional performance, such as audit, can provide useful information for those considering using such methods, but cannot replace judgement. In particular, there is no single phenomenon that can be labelled 'the Hawthorne effect'. The process of triangulation, considering a subject from different perspectives, might overcome the problems of Hawthorne effects better than using a single method such as controlled trials.