This study proposes a simple classification method for drivers based on their driving confidence and authority using feedback from steering wheel angle and traction motor current measurement data. Data is collected from 38 drivers on a prototype electric vehicle. For the classification process, the driving performance of each driver is assessed with respect to an expert driver. The data is time-stamped and geo-referenced. The experimental driving data is comprised of the two command inputs that the driver generates to regulate handling and lane-level tracking. Deviations from the expert's performance data are considered to be a measure of potential failure of the driver's confidence and authority leading to a possible accident. The dataset of the participating drivers is processed by extracting the mean, standard deviation, skewness and kurtosis values. Extracted features are evaluated with different classification methods, including KNN.