Multimodal Dataset of Smartphone Touchscreen Kinetics and Inertial Sensor Logs for Mindful and Mindless Scrolling

This dataset provides high-frequency multimodal behavioral logs captured during smartphone usage, specifically focusing on the distinction between mindful and mindless scrolling behaviors on social media. The data was collected to facilitate research in Human-Computer Interaction (HCI), behavioral biometrics, and digital wellbeing. The dataset comprises logs from 20 participants. To ensure hardware consistency and eliminate sensor variance across different devices, all data was recorded using a standardized research smartphone. Interactions were captured using a custom-built logging application. Participants performed two distinct sessions: a "mindful" session where they were given specific search tasks, and a "mindless" session involving free-form, dissociative scrolling through a TikTok feed. The repository includes: 1. Touchscreen Interaction Logs: Detailed records of interaction events (CLICK and SCROLL) including relative scroll distances (ScrollDeltaX and ScrollDeltaY) and the active application package name. 2. High-Frequency Inertial Sensor Data: Synchronized 3-axis accelerometer and gyroscope logs captured at a combined sampling rate of approximately 105 Hz, providing precise physical orientation and movement data. 3. Participant Metadata: Anonymized demographic information, including age, gender, and handedness (dominant hand). 4. Data Dictionary: A comprehensive guide explaining the schema, units, and column headers for all provided CSV files. This dataset is suitable for researchers interested in: - Developing machine learning models to detect cognitive states or "mindless" behavior in real-time. - Analyzing behavioral biometric patterns through scroll deltas and sensor kinetics. - Studying the physical kinetics of social media interaction on high-frequency data.