How can a sample can give information about bad products in whole lot?

[ali2233]

i have lot N = 10, i know nothing about the quality of the whole lot, i took sample of n = 5 , I found 2 bad products.

it is easy to get probability value for choosing 2 bad products using hyper geometric distribution, but what if i know nothing about whole lot.

Does this give any information about the distribution of bad products in the whole lot. ? does sample value give any information of bad ones in the whole lot.

is there any mathematical expression for this ?

 

[ali2233]

Re: how does sample can give information about bad products in whole lot

I really hope some one could share his ideas ,,,

 

[Marc]

Re: how does sample can give information about bad products in whole lot

It is a holiday weekend in the US and things are slow here so please do not expect an answer right away.

I'd try to help but I'm definitely not a math person.

 

[Bev D]

you can calculate the confidence intervals about 2/5 defects. There are several formulas and approaches (exact Binomial or Clopper-Pearson). But with such a small lot size, you are safest in saying that the lot may have as low a defect rate as 20% (you found the only 2 defects in the lot) or as high as 70% (you found the only 3 good units).

Ask yourself: Will either answer change the action you will take compared to the sample defect rate of 40%? Any confidence intervals you calculate will have a similar spread due to the 'small' sample size. If you need a really precise answer, you will need to inspect all units in the lot.