How will classroom simulators change teacher training?

March 17, 2023
Sam Sims.png

Sam Sims

Research Lead, Ambition Institute

For decades, pilots, doctors and astronauts have used simulations of reality to train in high-stakes activities. Our Research Lead, Dr Sam Sims, explores the potential for digital simulators to support professional development for teachers and leaders.

Ambition Institute’s new Research team just published our first study. We asked 89 trainee teachers to lead some retrieval practice, asking questions about sound to a group of primary school students. After their first attempt, we gave some of the trainees access to a video model exemplifying effective retrieval practice, while the others were given a written evidence summary on the same topic. All the trainees then went back into the classroom and repeated the retrieval exercise.

On the surface, this sounds like a normal study. However, there was no classroom and there were no pupils. Our study took place entirely within a simulated classroom environment. The only thing that was ‘real’ were the trainee teachers that participated and the evidence on retrieval practice.

Simulations are models that mimic the operation of real life. They are particularly useful for supporting practice in otherwise high-stakes settings and have long been used to help develop expertise. Pilots have been training in simulators since 1934, medics have been resuscitating mannequins since 1960, and NASA has been training astronauts in virtual reality since 1989. Beyond simple role plays, simulators have taken longer to arrive in education. However, this has now begun to change, with the introduction of classroom simulators such as TeachME and TLE TeachLivE.

We used the Mursion simulator in our research. This is a mixed reality simulator, combining automated elements controlled by a computer with elements controlled by a human ‘puppeteer’, or simulation specialist. The simulation consists of five pupil avatars sat at a desk in a classroom, who can be made to speak and perform certain actions, such as looking disappointed. The five pupils can also do certain thing in unison, such as raising their hands.

"Anyone who doubts that computers can realistically simulate a student should ask ChatGPT to ‘explain the causes of World War 2 in the style of 13-year-old."

I trialled Mursion before we began our research and was pleasantly surprised by how immersive it felt. Participants in our study seemed to feel the same way. Just over 94% of participants reported that they enjoyed taking part in the simulator. One participant said “I think working with a simulator is a great idea to motivate and build trainee confidence”. Another commented that “This has been very helpful…. A great way to practise retrieval outside the pressure of being in the classroom as a trainee.”

So how will classroom simulators change teacher professional development? Existing mixed reality environments are not an economical approach to at-scale teacher training because each trainee requires input from both a simulation specialist (who manages the simulated environment) and a teacher educator. However, recent rapid advances in technology mean this is likely to change. Anyone who doubts that computers can realistically simulate a student should ask ChatGPT to ‘explain the causes of World War 2 in the style of 13-year-old’. Anyone who doubts that computers can give useful feedback should read this paper in which researchers used a large language model (similar to ChatGPT) to help trainee teachers diagnose student learning difficulties.

For the time being, however, classroom simulators are proving particularly useful for conducting research on how to design teacher professional development. Simulators allow teachers to participate in research from anywhere that they can plug in a laptop, thus cutting the costs of research. Simulated environments also provide researchers with experimental control that would otherwise be hard to achieve, which allows them to address new types of research questions.

Useful findings from simulator studies are now beginning to pile up. Pioneering work by Julie Cohen and colleagues has shown that instructional coaching is more effective than self-reflection in improving teachers’ classroom management, and their ability to discuss texts with pupils. In our new research, we found that adding a video model to some professional development dramatically improved trainee teachers’ use of questioning for retrieval.


Of course, not all questions can be addressed in simulators. Some of these limitations relate to the simulation tools themselves. For example, Mursion does not allow for students to respond to teacher questions on mini-whiteboards, or for teachers to circulate around the room. However, improvements in technology may relax some of these constraints in the coming years.

Other limitations of simulators are more baked in. For example, it is hard to see how we will ever be able to study the effect of different approaches to teaching on pupil learning, when the pupils in question are merely avatars. That’s why we focused our simulator study on developing teachers’ skills in using a pedagogical approach for which there is already considerable evidence of real-world impact on pupil learning.

Despite these limitations, I predict that we will learn a lot about how to design teacher professional development from simulator studies in the next few years. If you’re interested in learning more about our new simulator study, you can find the full research report here and a three page non-technical summary here. Watch this space for more research, conducted both in the real world and in the metaverse.

Search blog posts by topic: