Because simulators offer the possibility of functioning as authentic representations of real-world tasks, these tools are regarded as efficient for developing expertise. The users’ experience of realism is recognised as crucial, and is often regarded as an effect of the similarity between reality and the simulator itself. In this study, it is argued that simulation as a realistic and relevant activity cannot be predesigned but emerges in the interaction between the participants, the simulator, and the context. The study draws on interaction analysis of video data from medical training. The aim is to contrast the use of two different simulators to explore the requirements needed to establish and maintain simulations as authentic representations of clinical practice. Irrespective of the realism of the simulator, glitches in the understanding of the simulation as work-related activity appear and are bridged by participants. This regularly involves an orientation to the relevant similarities with work and, simultaneously, the ruling out of irrelevant dissimilarities. In doing so, the participants rely on established professional practices to construe the situation. Moreover, the realism of the simulation is maintained through the participants’ mutual orientation to the moral order of good clinical practice and a proper simulation. It is concluded that the design of simulation activities needs to account for the possibilities of participants understanding the specific conditions of the simulation and the work practices that the simulation represents. Learning to simulate is thus something that needs further attention in its own right.
