Construction of a model for discriminating between electroencephalographic patterns at the time of incorrect inputs based on sensitivity spectrum analysis

Abstract

We consider the possibility of prior discovery of slip-type human error by quantifying the "distraction" state that precedes it. We investigated brain activity during the task of keyboard-based text input taking incorrect input as a slip-type error. We used sensitivity spectrum analysis to quantify four output states. We attempted to detect in correct input by discriminating between the output of the distraction state before and after incorrect input and that of the state before correct input. © 2011 Springer-Verlag.