Calculating Adequate Receiving Inspection Sample Size

[Mayank Trivedi]

Hi All,

We process incoming work from a different country and a random sample is taken out of the lot for Quality Check.

We need to understand if:
1) Sample size is adequate and if not what should be the correct sample size. Also if we go with the present sample size what is the margin of error that we are looking at to meet the quality target of 99.10% month on month?
2) Also is the sampling methodology biased in favour of those who process fewer jobs, or towards any one location?

I have posted the data for 2 months.

Accuracy scores are calculated on a point based system. There could be certain errors which are dearer than others and hence with equal number of errors the scores of two people will be different.

Regards,

 

[Stijloor]

A Quick Bump!

Can someone help Mayank?

Thank you very much!!

 

[Steve Prevette]

We need to understand if:
1) Sample size is adequate and if not what should be the correct sample size. Also if we go with the present sample size what is the margin of error that we are looking at to meet the quality target of 99.10% month on month?

Response: With thousands of jobs being checked, I'd say you have plenty for sample size. Sample size requirements are NOT directly related to the population. Above a population size of a few hundred, the binomial distribution takes over, which assumes an infinite population size.

The determination of the "Correct" sample size is made more difficult with your scoring scheme, but I'd say just a cursory overview I would accept it is an adequate sample size.

2) Also is the sampling methodology biased in favour of those who process fewer jobs, or towards any one location?

Response: As long as every individual job that is processed has an equally likely chance of being chosen, then the size of jobs and location activity levels are self-correcting.