Aviation selection test battery component predictiveness of primary flight training outcomes among diverse groups

Abstract

The Aviation Selection Test Battery (ASTB) has been the qualifying benchmark for the Naval Aviation since World War II. While it is necessary that test scores effectively select the candidates with the greatest chance for success, the ASTB strides toward increasing diversity while maintaining low attrition. Using archived Student Naval Aviator and Student Naval Flight Officer ASTB subtest scores and Primary Flight Training (PFT) records, this study examined the ASTB's predictive ability with respect performance in PFT. Specifically the study consists of two analyses: 1) determine how well the ASTB could predict majority and minority group performance in primary flight training; and 2) determine how well the ASTB could predict success in each training phase and for the entire sample and select groups. The linear regression analysis successfully fit a significant model for the entire sample and Caucasians, but was unable to produce a significant model for African Americans or Hispanics, as there was insufficient data available for either group. The model, when fitted to the entire dataset, with race as an independent variable, yielded a result where all independent variables were significant. The results from the logistic regression models showed there was evidence that four of the ASTB subtests were significant and positive predictors for the entire sample and Caucasians; but was unable to produce a significant model for African Americans or Hispanics. It is apparent that the small data set for minorities limited this study. Efforts to collect data from personnel records should be conducted to obtain all scores from flight training, so that these groups can be further investigated.