STREX-VR (Stress Tracking EXperience in Virtual Reality)

This dataset was collected from participants playing a first-person VR horror survival game set on a deserted space station. The game was designed to induce stress through environmental challenges, resource scarcity, and threatening enemies. Data includes physiological, behavioral, and self-reported stress measurements. Physiological signals were recorded using the Empatica E4 wristband, capturing Electrodermal Activity (EDA) at 4 Hz, a physiological indicator of arousal. Behavioral motion data were gathered from the Oculus Quest 2 headset and controllers, recording position, rotation, velocity, and acceleration at 64 Hz. In-game actions, such as button presses and pressure intensity, were also logged. Before the gaming session, participants completed two questionnaires: Perceived Stress Scale (PSS) questionnaire: A standard phsycological 10-item questionnaire rated on a 5-point Likert scale (0-4) to assess perceived stress over the past month. Demographic questionnaire (accessible at https://forms.gle/i4syZb2CpXd91r9M8): Collected unchanging personal data, including age, gender, gaming habits, and VR experience. After gameplay, participants reviewed their recordings and annotated perceived stress levels using DANTE (Dimensional ANnotation Tool for Emotions), a web-based tool that enables continuous stress assessment on a scale from -1 to 1 (step of 0.001). A color-coded bar indicated low, medium, and high stress levels. The game’s structured design ensured consistent exposure to stress-inducing stimuli across all participants, enabling reliable analysis of physiological and behavioral responses in an immersive VR environment. Here is the description of the files, with each row corresponding to a participant: stress.csv contains the following columns: TimeStamp: A time counter starting from 0 for each session. UserId: A unique identifier for the participant. NameVideo: The name of the video associated with the recorded data. AnnoType: The type of annotation (e.g., stress, valence, arousal). Value: The recorded stress level at the corresponding moment. EDA.csv includes: First row: An initial timestamp in numeric format, representing Unix epoch time. Second row: The sampling rate value (in Hz). Remaining rows: EDA signal values recorded over time.. S#.csv (files containing motion behavioral data collected from Oculus) include: Movement data for the head and hands: velocity, angular velocity, acceleration, and angular acceleration. Buttons data: Grip pressure and trigger pressure for both hands. A column TimeStamp in YYYY-MM-DD HH:MM:SS format. PerceivedStressScale.csv: Contains responses to the PSS questionnaire. EarlyQuestionnaire.csv: Includes demographic data. If you use this dataset, please cite the following paper: Stress Assessment in Virtual Reality Horror Games Using Players' Behavioural and Physiological Data, 2025 (Forthcoming). DOI: 10.1109/TG.2025.3563258