In failure rate modelling, is it better to use single quality and environment factor or seperate set of environment factors for individual resistors?
