Neural and physiological dynamics before and after unpredictable collision events in virtual reality using novel wearable sensing

This project examines whether pre-stimulus and post-stimulus physiological activity differ between unpredictable simulated car collision and neutral (no collision) events. While the primary focus is on predictive anticipatory activity (PAA; changes in physiological signals that occur prior to randomly selected events in the absence of explicit predictive cues), the realistic, immersive nature of this VR paradigm also enables investigation of post-stimulus responses to collision events. The study aims to carefully control for potential confounds that could produce similar anticipatory patterns (e.g., order effects, expectation bias, time-on-task effects). Participants will take part in a virtual reality (VR) driving simulation presented using a head-mounted display. They will be seated in a passive driving scenario in which they observe a road while their own vehicle moves forward automatically. On each trial, a car appears in the opposite lane. In one condition, the car simply passes by without incident (no-collision trials). In another condition, the car unpredictably crashes into the participant's vehicle (collision trials). Importantly, participants receive no sensory or temporal cues indicating whether an accident will occur on a given trial. Crash and no-crash trials are randomly interspersed, and trial order will be determined using a quantum random number generator (qRNG). A fixed inter-stimulus interval (ISI) will be used to control for temporal expectation effects, as variable ISIs elicit CNV activity related to predicting when events occur, which would confound the detection of anticipatory activity related to what type of event will occur. The study employs a fully passive paradigm, with no required motor response, to avoid confounds related to movement preparation. Physiological data will be recorded continuously using the novel Galea system (by OpenBCI, Inc.), which integrates Varjo Aero virtual reality (VR), electroencephalography (EEG), electrodermal activity (EDA), photoplethysmography (PPG; heart rate), and inertial measurement units (IMU) within a single wearable platform. Analyses will examine both the pre-stimulus period preceding each event and the post-stimulus period following each event, comparing physiological activity surrounding collision versus no-collision trials. This temporal approach allows examination of anticipatory signatures before events and reactive responses after events. The project has five objectives: (1) Discriminate the post-stimulus physiological correlates between unpredictable collisions compared to no-collision events. (2) Investigate whether pre-stimulus physiological activity differs before unpredictable collisions compared to no-collision events. (3) If anticipatory effects are observed, determine whether they are related to or modulated by: (a) the contingent negative variation (CNV), a well-characterized slow negative EEG potential in frontocentral regions associated with temporal expectation; (b) expectation biases arising from perceived patterns in random sequences; (c) time-on-task effects such as habituation; and (d) gambler's fallacy-type reasoning where participants may implicitly expect event probabilities to balance over short timescales by learning the stimulus design. (4) Examine whether anticipatory and reactive physiological effects are modulated by third variables such as age, gender, and prior driving or athletic experience. (5) Explore whether combining multiple physiological modalities using machine-learning approaches improves the detection of anticipatory effects compared to single-modality analyses. All hypotheses, preprocessing steps, and statistical analyses will be preregistered. Anonymized data and analysis code will be made openly available. The results are expected to clarify whether anticipatory physiological activity preceding unpredictable events can be robustly observed in an ecologically valid, immersive environment, and whether such effects can be distinguished from conventional anticipatory potentials and cognitive biases. This study employs state-of-the-art analysis methods and addresses key methodological limitations identified in prior research.