Hello to all the gurus!
I’ve been studying B vs. C, which I believe is a Shainin technique, and would like to apply it to reliability validation. I’m not quite sure I have it right though.
I have a current product – the C – that has been tested in the past with much cost to be 95R/90C at 100K cycles. If I perform B vs C and it ends up that I’m (for example 99% confident) that my B product is better than the C – can I say that with 99% confidence my C product is 95R/90C at 100K cycles?? Or have I mixed this up.
We have spent an enormous amount of $ to perform reliability testing on the C product. The goal is take advantage of what we do know (the reliability) and without much more testing prove that the B product meets the reliability target.
~thanks in advance!
First of all there are no gurus here. Those that try to put themselves out there as such are just bags of hot air.
If your company has spent tremendous resources already, you should consider also spending a little on understand real reliability stats, not the Shainin hocus pocus.
Shainin’s methods are valid under their assumptions but they never give insight to those assumptions. Learning reliability would make you more versatile and you would understand your data better. Reliability stats are well documented and not hard to understand. They are also incorporated into all legit stat packages (those that are not just excel addins)