A phenomenology of direct observation in residency Is Miller's 'does' level observable?

Our results indicate that the 'observer effect' is much more material than was previously understood. Consequently, observing residents' 'authentic' behaviour at Miller's 'does' level, as if the supervisor were not there, seems impossible and a misleading concept: misleading, because it may frustrate residents, and cause supervisors to neglect patients' and residents' needs in DO situations. We suggest that one-way DO is better replaced by bi-directional DO in working-and-learning-together sessions.