When and How to apply the Bayes Success Run approach for Sample Size Determination

[v9991]

when, how do we apply Bayes Success run approach for sample size determination.

my layman's approach, is that its applied in the context of 'reliability' based assessment.!

 

[Statistical Steven]

Re: When and How do we apply Bayes Success Run approach for Sample Size Determination

You are correct. In layman's terms it is a statement about the reliability of the population based on a sample where you find no defectives in the sample. as the run of successes increases you have more reliability in the population that it has a low defect rate. When used apriori, you can say if I sample 59 units and find no defects, I have 95% confidence with 95% reliability about my process (or my population from which I sampled). This is usually applied to attribute sampling and not variable sampling plans.

 

[Bev D]

Re: When and How do we apply Bayes Success Run approach for Sample Size Determination

when, how do we apply Bayes Success run approach for sample size determination.

my layman's approach, is that its applied in the context of 'reliability' based assessment.!

We must be careful as statisticians take normal words and give them 'special' meanings.

I'm not sure what you mean by "reliability assessment". If you are referring to reliability engineering context than I would then you are misinterpreting the use of the term in this sampling scheme. The two definitions are very different. This sampling scheme Can be useful for many sampling situations and not just Reliability engineering applications

 

[v9991]

Re: When and How to apply the Bayes Success Run approach for Sample Size Determinatio

Agree, I am using the 'reliability' in appropriate context, i am not mixing up with the reliability analysis related dimension.


can you pl suggest recommended approach for determining sample size in following scenario (s)

1. a new product, first production batch (verification / validation) - with little or limited assessment on variability.
( my guess, is that bayes success run could be reasonable beginning for determining sample size.)

2. post validation, routine commercial batch. (acceptance sampling for a given (my guess, is to switch over to tolerance interval based assessment.)

3. an investigation on a batch, for which routine analysis revealed certain failures/outliers. ( my guess is to take up a confidence interval based approach., ensuring power - differnce - variability, minitab option of sample size estimation.)

thank you all in advance for your guidance and clarification

 

[v9991]

Re: When and How to apply the Bayes Success Run approach for Sample Size Determinatio

Agree, I am using the 'reliability' in appropriate context, i am not mixing up with the reliability analysis related dimension.


can you pl suggest recommended approach for determining sample size in following scenario (s)

1. a new product, first production batch (verification / validation) - with little or limited assessment on variability.
( my guess, is that bayes success run could be reasonable beginning for determining sample size.)

2. post validation, routine commercial batch. (acceptance sampling for a given (my guess, is to switch over to tolerance interval based assessment.)

3. an investigation on a batch, for which routine analysis revealed certain failures/outliers. ( my guess is to take up a confidence interval based approach., ensuring power - differnce - variability, minitab option of sample size estimation.)

thank you all in advance for your guidance and clarification

actually, we are in the process of upgrading the sampling procedure and strategy for validation sampling approach; and above scenarios were something we were attempting to describe the approach.