HRV biofeedback training in an active Virtual Reality decision-making game: Can it be trained?

The present preregistration describes a study for which data was collected between April and June 2021. The data has not been analyzed yet. The overarching goal of this study was to test the effectiveness of a VR-based biofeedback (BFB) training for police officers. The training aims to improve control of Heart-Rate variability (HRV) in an active and immersive VR context, in which police officers make shooting decisions under threat. The unique novel aspect of the study is that BFB training is integrated in an active decision-making context. Our first aim is to validate whether the VR training environment is indeed arousal-inducing and evokes the expected in-game behavior. To this end, we will compare the VR environment and a baseline assessment with respect to Heart-Rate (HR) changes and self-reported stress responses. Additionally, we will investigate the effect of behavioral priming on in-game shoot-don’t shoot decisions. The second aim is to investigate the effectiveness of the training. Therefore, we compare an experimental group (who received the training) with a control group (who did not receive the training but played the VR game once) with respect to in-game psychophysiological, behavioral and self-report data. We additionally investigated if biofeedback delivered in action was the causal mechanism explaining an improvement by changing the order in which participants in the experimental group receive the first biofeedback session in VR. Last for this aim, we investigate if the training has an effect on the feeling of self-efficacy and physiological awareness. Our third aim was to assess whether acquired HRV regulation skills would transfer to a police-relevant independent arousing decision-making context, outside of the game. To this end the experimental and control groups were compared with respect to HRV and shoot/don’t shoot decisions during a simulation exercise inspired by the dispatch priming task of Taylor (2020) that served as an independent transfer task, administered after the final game session. Below we only give an overview of the design and outcome measures of the current dataset, for details on the rationale and methods see previous reports (Brammer et al., 2021; Michela et al., 2022).