Product data reliability

[tryinall]

I am looking for the best way to determine the amount of inconsitency in our shop product data. ie. I am thinking of checking 100 products that will either pass or fail the test of number of units on the shelf is equal to the number "on hand" in our inventory system. The total population is 28000 products. This test is to measure the accuracy of our product data. What would be the best method/statistical application? Once I have the numbers on the sample set, how do I use it to estimate for the population? I used to have a grasp on this stuff and am looking to come back to speed. Cheers!

 

[BigBeachBanana]

I am looking for the best way to determine the amount of inconsitency in our shop product data. ie. I am thinking of checking 100 products that will either pass or fail the test of number of units on the shelf is equal to the number "on hand" in our inventory system. The total population is 28000 products. This test is to measure the accuracy of our product data. What would be the best method/statistical application? Once I have the numbers on the sample set, how do I use it to estimate for the population? I used to have a grasp on this stuff and am looking to come back to speed. Cheers!

You think you're trying to do hypothesis testing. If so, you need to state your null hypothesis e.g. less than 50% of the products will pass.

 

[tryinall]

Many thanks, @BigBeachBanana. Say, if my null hypothesis were H(0) - less than 80% of the product will pass. Am I right in assuming that I must next specify my confidence interval and error? Typically 95% and 3% respectively?

 

[BigBeachBanana]

Many thanks, @BigBeachBanana. Say, if my null hypothesis were H(0) - less than 80% of the product will pass. Am I right in assuming that I must next specify my confidence interval and error? Typically 95% and 3% respectively?

A 95% confidence interval corresponds to a 5% error.
Confidence interval + error = 1.