Training in the CV90 simulator

Two studies at the CV90 simulator (Sw. Befattningstränare Stridsfordon 90, BFT 90) have been performed. The first study was performed in the autumn 2009, during the early part of the soldiers' education. The second study was performed during the later part of the soldiers' education. 2009 only gunners participated, whereas drivers and commanders were included 2010. The focus of the studies were fidelity, presence, fun, motivation, feedback, learning, and transfer of training. The results show that the gunners rated learning in the simulator and the trainings effect on performance in the real CV90 higher than the other positions. The gunners had also trained more in the simulator. During early education 2009, the gunners trained more on basic handling and preparation before training in the real CV90, compared to later when they trained more on team training. Learning in the simulator was rated higher by the gunners during early education compared to later, for the other studied training concepts there were only small differences. It seems reasonable that learning in the simulator is rated higher in early stages with much procedural training in the simulator, compared to later stages with a larger focus on team training in the real CV90. In spite that the participants had many negative views on the graphics and fidelity in the simulator, their ratings of how much they learnt during training and the effect of the training on performance in the real CV90 were relatively high. This indicates the possibilities of training effects in simple low fidelity simulators. By component analysis and linear structural equation modelling (LISREL) a model of the casual relations between the studied training concepts could be created. The model shows that the training environment (measured by how fun the training was experienced, motivation, feedback, presence, and fidelity) had a positive effect on the training in the simulator (measured by learning in the simulator and its effect on performance in the real CV90. Training in the simulator could also be split into a two components, which led to a model with three components where the training environment has a positive effect on the simulator training which in turn has a positive influence on transfer of training to the real CV90. The results can also be used for instrument development, concerning the questionnaire used at the studies. To provide recommendations on improvements concerning training and simulator development, studies with a more qualitative approach are needed.