Hello again, I hope you remembered I mentioned box sizes in the last post.

In usual quality “lingo” the box size is called lot size, so it is not the physical size of the box, but the amount of products within the box I was referring to.

In ISO2859 (and many other standards written about sample plans for inspection) lot size is one of the factors that affect how many samples are needed to make sure the shipment is acceptable.

However; in “small print” you can read this is not related to the actual probability of accepting poor quality (or vice versa; scrapping good parts). The chance of finding one red ball is 10% no matter if the lot size is 1,000 or if it is 1,000,000 (see last post about the 10%).

Why then increase the sample with increasing lot size?

If the parts are produced over a period of time, and stored in many different bags, boxes and pallets, then you should make sure you did not only inspect the last box made Friday afternoon.

If you have many parts ready for delivery a mistake could be very costly compared to a small shipment and should be inspected more carefully.

These two reasons are the main reasons for increasing sample sizes for larger lot sizes.

Inversely this also could (should?) be interpreted this way; If you can ensure a homogenous sample even with large lot size and if you can afford a mistake: Use a risk based approach to decide if a smaller sample is acceptable (but make sure you justify and document this).

One fully “legal” way to reduce the sample size is to apply another inspection level than the typical level II (ISO 2859-1).

This way it is not “reduced inspection” and you will still follow the standards advice to consider lot size and in most situations the consumer risk does not increase significantly.

So next time you need inspection – give it some thought, it may save you some time.

Good luck with your sampling

For more information please contact EpsilonPlus here.

To read more interesting posts have a look at this page.