This paper presents the results of a survey conducted by the G41 Reliability Committee of the Electronic Industries Association (EIA). Also included in a closely related study of an ARINC Research Corporation investigation of the relationship between acceptance-test reliability and operational reliability.
The basic statistical technique used in the ARINC Research study was multiple regression analysis. In investigating the relationships, a number of factors were used other than the inherent hardware reliability (as indicated by the results of AGREE tests). These factors included: equipment descriptors, data collection methods, and a combined environmental mission and application descriptor. These factors were considered as a means of increasing precision in estimating expected field operational results.
The data received as a result of the survey has been summarized in tabular form, but falls into many classifications due to the variety of equipment types, the wide range of missions of the equipment, and the variation in estimating and testing. The results were not very suitable for detailed analysis, due to the small size of the samples available.