Examples of Aircraft Scheduled-Maintenance Analysis Problems

A description of several statistical analyses of aircraft scheduled maintenance data illustrating the kinds of errors that might occur through the use of packaged computer programs and procedures by nonspecialists. The analyses are grouped to illustrate three potential sources of analytical error: (1) alternative causes for observed phenomena, (2) variations in sample size, and (3) failure to compensate for extraneous effects. The first is illustrated by analyses assessing the before and after depot inspection condition of the F-4C and F-4D aircraft and F-106 performance as a function of the time between inspections; the second, by an experiment to determine the effects of base-level inspection on the reliability of J-57 engines in the F-100; and the third, by a study to determine whether depot maintenance affects the likelihood of mishaps on the C-141A and an analysis to determine the effect of extending the time between inspections of the C-141A. In each case skillful analysis either reversed or nullified the original conclusions.