Laboratory performance prediction using virtual reality behaviometrics.

Area of Science:

• Biometrics
• Virtual Reality (VR) in Education
• Laboratory Skills Assessment

Background:

• Traditional laboratory skills assessments can be time-consuming and subjective.
• There is a need for objective, reliable, and efficient methods to evaluate practical skills in educational and biopharmaceutical settings.

Purpose of the Study:

• To investigate the efficacy of virtual reality (VR) behaviometrics for assessing compliance and physical laboratory skills.
• To identify behavioral predictors of expertise in laboratory tasks using VR.

Main Methods:

• Utilized machine learning and classical statistics, including logistic regression, to analyze behavioral data from VR simulations.
• Developed VR tasks simulating practical laboratory skills, such as pH meter handling.
• Collected data on game scores and interaction frequency within the VR environment.

Main Results:

• A logistic regression model achieved 77% accuracy in classifying individuals as experts or novices in pH meter handling.
• Key performance predictors identified include game score and the number of interactions within VR tasks.
• VR behaviometrics demonstrated a significant correlation with practical laboratory skill proficiency.

Conclusions:

• Virtual reality (VR) behaviometrics offer a viable, automated, and dependable method for assessing laboratory skills and compliance.
• This approach provides biopharmaceutical companies and academic institutions with an efficient alternative to traditional in-person evaluations.
• Integrating performance assessment directly into VR training tools can eliminate the need for separate, laborious post-training assessments.