Prediction and Classification of Operational Errors and Routine Operations Using Sector Characteristics Variables

Abstract

This study examined prediction and classification of operational errors (OEs) and routine operations (ROs) using sector characteristics variables. Average Control Duration, Aircraft Mix Index, Average Lateral Distance, Average Vertical Distance, Number of Handoffs, Number of Point Outs, Number of Transitioning Aircraft, and Number of Heading Changes were used as predictors in two stepwise logistic regression analyses conducted for the high-altitude and low-altitude sectors. In the high-altitude sample, variables included in the final model (Number of Heading Changes, Number of Transitioning Aircraft, and Average Control Duration) accurately classified OE and RO samples for 80% of the cases. In the low-altitude sample, variables included in the final model (Number of Point Outs, the Number of Handoffs, and the Number of Heading Changes) accurately classified OE and RO samples for 79% of the cases. Although logistic regression cannot be used to determine causation, it effectively identified variables that predicted the occurrence of OEs.17.