Probability of detecting a failure (binomial distribution)

Hi All,

I have a real life problem that I need to solve (not homework). If anyone could take a look and see if I am approaching this problem correctly I will greatly appreciate your help.

I need to predict the percentage of bad die I will not detect using the following scenario.

I know that my average failure die failure rate is 7%
I know my failure rate varies from wafer to wafer
6 Die are tested per wafer, if 2 or more die fail, the entire wafer is rejected.

I've used the binomial probablity formula to predict the probabiliy that a wafer will pass dependent upon the % of bad die per wafer. Can I assume that my average failure rate will be equal to the average failure rate of the population? Or, do I also need to know the distribution of die failure rates on the individual wafers (See column A of attached spreadsheet)?

Please let me know if I am approaching the problem correctly or if I should be using a different method.


Last edited: