Test Confidence

[Robbie]

If I design a test (e.g. no of test factory cycles to demonstrate MTTR), what advice is offered on how to select a desired confidence. Why prefer 95% instead of 75% (apart from the obvious). Would it come to practicalities and constraints around what was possible to test e.g. only have the capacity to carry out x number of cycle tests, and will therefore get the confidence associated with it?

[David]

Hi Robbie,

I believe that you are referring to the confidence level when trying to develop a demonstration test. In terms of the confidence level, it really comes down to how much risk you are willing to accept. As you increase the confidence level, the required number of units/test time is going to also increase. For a medical device, a higher confidence level will generally be used since you do not want to have much risk. For something less critical, you can use a lower confidence level.

I hope this helps.

[Robbie]

Thanks David