Simulation Training That Works – the Research Behind It
Pilots train in flight simulators. Surgeons practice on models before operating on people. But managers and project managers are often expected to learn difficult conversations and decisions live, with real people and real consequences.
Researcher Eduardo Salas, one of the world's most cited scholars on training and teamwork, examined together with colleagues Wildman and Piccolo what it takes for simulation-based training to actually work in leadership and management education. Their 2009 article remains one of the clearest guides for anyone who wants to train soft skills seriously.
What does the research say?
Salas and his colleagues reviewed what the research knew about simulations in management education and reached a conclusion that surprises many: the simulation itself is not what creates the learning. An engaging exercise can feel instructive without changing anything about how you act afterwards. What matters is how the training is designed around the simulation.
Two things stand out as decisive. The first is realistic, active practice: the participant must make the decisions and hold the conversations themselves, not watch someone else do it, and the situation must resemble reality closely enough for the behaviors to be the same. The second is a structured review afterwards, a so-called debrief, in which the participant, with support, gets to see what they actually did, why it had the effects it had, and what could be done differently.
Without the structured debrief, the simulation risks becoming entertainment: fun in the moment, forgotten the following week. With it, the simulation becomes one of the most effective forms of training we know of, because it combines action, emotion, and reflection in a single process. The researchers' message is therefore not to simulate more, but to simulate right: practice plus feedback plus reflection.
What does this mean for you as a project manager?
Set demands on your own training. If you attend a course with role-plays or exercises, ask how the debrief afterwards is done. Do you get concrete feedback on what you said and did, or just a generic "good job"? The debrief is where the learning happens, so that's what you should judge the course by.
Use the principle in your daily work. After an important negotiation or a difficult delivery message: run your own mini-debrief. What was my goal? What did I do? What effect did it have? What will I try next time? Ten minutes right afterwards is worth more than an hour of brooding a week later.
Train your team the same way. When the team rehearses for an important client presentation or a crisis, spend at least as much time on the debrief as on the exercise itself. Just running through it is practicing on the surface; reviewing it afterwards is learning in depth. Consider appointing someone whose only job is to observe during the exercise, so the debrief can be based on what actually happened instead of what the participants remember.
How to practice this
This is exactly the research that determines how a good conversation simulator should be built: the conversation with an AI counterpart is only half the training, and the evidence-based debrief afterwards, where the feedback points to your own words in the transcript, is the other half. If you want to experience the combination of realistic practice and a structured debrief, you can try Project-simulator free for a week.
In Project-simulator you practice these conversations against an AI counterpart and get feedback grounded in research like this.
Try free for 7 days