Yeah that’s why I don’t like the use of ‘probability’ when it comes to determining the occurrence level when discussing product quality risk determinations. Probability and occurrence rate - or frequency - while similar do not share the exact same DNA. Frequency of occurrence (aka defect rate) can be determined by testing and should be. (This is done throughout development as we determine specifications and then during final testing by testing at teh extremes of the specs and use conditions. Actual data not biased guesses and wishful thinking).
Of course in the human plane of existence there is no such thing as a truly random event. Random means without cause. There is always a cause. In statistics we say data are random when they appear Randall - without any apparent cause or any discernible pattern. This is the ‘common cause’ variation referred to by Shewhart. And of course there is the ‘random’ or ‘out of the blue, unexpected accident or event that throws a spanner into the works - in FMEA we are supposed to anticipate these types of events and mitigate their effect - not their occurrence: in aerospace it may be a bird strike in the engine, in automotive it may be icy roads that cause a wreck…
Of course in the human plane of existence there is no such thing as a truly random event. Random means without cause. There is always a cause. In statistics we say data are random when they appear Randall - without any apparent cause or any discernible pattern. This is the ‘common cause’ variation referred to by Shewhart. And of course there is the ‘random’ or ‘out of the blue, unexpected accident or event that throws a spanner into the works - in FMEA we are supposed to anticipate these types of events and mitigate their effect - not their occurrence: in aerospace it may be a bird strike in the engine, in automotive it may be icy roads that cause a wreck…
