Investigation of EEG-based indicators of skill acquisition as novice participants practice a lifeboat manoeuvering task in a simulator

Abstract

Adequate training is essential in safety critical occupations. Task proficiency is typically assessed through relevant performance measures. While such measures provide information about how effectively an individual can perform the task, they give no insight about their comfort level. Ideally, individuals would be capable of executing tasks not just at a certain level of performance, but also with confidence and a high degree of cognitive efficiency. Neural signals may provide information regarding a trainee’s task proficiency that performance measures alone cannot. The purpose of this study was to investigate patterns in neural activity that are indicative of task proficiency. Ten novice participants completed ten trials of a manoeuvering task in a high-fidelity lifeboat simulator while their neural activity was recorded via 64-channel EEG. Power spectral features were used along with linear discriminant analysis to classify the data from pairs of consecutive trials. Repeated measures mixed model linear regression showed that on average, the classification accuracy of consecutive trials decreased significantly over the course of training (from 82% to 73%). Since the classification accuracies reflect how different the neural activation patterns in the brain are between the trials classified, this result indicates that with practice, the associated neural activity becomes more similar from trial to trial. We hypothesize that in the early stages of the practice session, the neural activity is quite distinct from trial to trial as the individual works to develop and refine a strategy for task execution, then as they settle on an effective strategy, their neural activity becomes more stable across trials, explaining the lower classification accuracy observed in consecutive trials later in the session. These results could be used to develop a neural indicator of task proficiency.