"What was he thinking?": Using EEG data to facilitate the interpretation of performance patterns

Abstract

Previous research has demonstrated that EEG data can be used to identify and remove unintentional responses from a data set (guesses and slips). This study sought to determine if removing this error variance has a significant impact on the interpretation of a trainee's performance. Participants were taught to recognize tank silhouettes. Multiple linear regression models were built for each participant based on three sets of their data: 1) all trials of their performance data, 2) only trials that were learned according to a state space analysis, and 3) their intentional data as identified by EEG. When compared to an expert model, each of the three models for every participant yielded a different diagnosis, indicating that filtering performance data with EEG data changes the interpretation of a participant's competence. © 2009 Springer.