Anonymized Student Event and Error Logs from a 3D Virtual Biotechnology Laboratory Procedural Training Environment

This dataset contains anonymized interaction logs collected from a 3D virtual laboratory used for procedural training in a biotechnology-style workflow. It is released to support open research in learning analytics, educational data mining, intelligent tutoring systems, and sequence modeling of student behavior. The deposition includes three semicolon-separated CSV files in UTC time: (1) StudentEvents.csv, an event stream of student actions and system events referenced by protocol keys; (2) Errors.csv, tutor-detected errors that distinguish between pedagogically relevant/non-blocking events and blocking errors (e.g., unmet dependencies, incompatibilities, repeated actions), including the event key associated with the error; and (3) ErrorsNoTutor.csv, low-level interaction errors detected without tutor intervention (e.g., spatial/interaction constraints with 3D objects). Each record is linked through an anonymized StudentId, enabling per-student reconstruction of action sequences and error trajectories. The dataset can be used to study procedural skill acquisition, error-driven learning, feedback design, and to benchmark predictive models of next action / error occurrence in complex simulated environments.