In general, the 217Plus/Prism models are based primarily on field reliability data. Therefore, failures are considered to have occurred when a part is replaced and corrects the specific symptoms that prompted the repair action. So, if the failure symptoms for the systems on which the field data was collected for the RIAC database was a manifestation of a data retention issue, then data retention failures would inherently be included in the models. However, for the majority of the data collected, the specific failure causes are unknown, and therefore we cannot be certain that they were included in every case.