Dear all,
I need some support in an implementation of random sampling at receiving inspection. During an ISO audit it was found that we do not have a proper randomization process in place and as a result there is a risk of sampling bias.
We have looked at ISO 24153 Random Sampling and randomization procedures. We have learned that there are various ways that we can use randomization procedures to assure that all products in a lot have approximately the same chance of being selected for inspection.
The challenge is how to implement this in practice. We receive goods in many different forms: boxed on pallets, trays in boxes of various sizes, loose parts in bags etc etc. We are now trying to understand how to map a list of random numbers onto the wide variety of packing methods.
Juran gives some suggestions (5th edition, 46.16) for systematically packed materials and for bulk packed materials but is not very explicit. My colleagues from stores are very concerned that we will end up with elaborate and complex procedures (me too).
I am certain that other Quality colleagues have resolved such issues in pragmatic ways. I would like to ask those people to comment on:
Paul
I need some support in an implementation of random sampling at receiving inspection. During an ISO audit it was found that we do not have a proper randomization process in place and as a result there is a risk of sampling bias.
We have looked at ISO 24153 Random Sampling and randomization procedures. We have learned that there are various ways that we can use randomization procedures to assure that all products in a lot have approximately the same chance of being selected for inspection.
The challenge is how to implement this in practice. We receive goods in many different forms: boxed on pallets, trays in boxes of various sizes, loose parts in bags etc etc. We are now trying to understand how to map a list of random numbers onto the wide variety of packing methods.
Juran gives some suggestions (5th edition, 46.16) for systematically packed materials and for bulk packed materials but is not very explicit. My colleagues from stores are very concerned that we will end up with elaborate and complex procedures (me too).
I am certain that other Quality colleagues have resolved such issues in pragmatic ways. I would like to ask those people to comment on:
- Do you have experience with random tables versus computerized random number generation?
- Can you give examples of how you map the random numbers onto specific packing methods (e.g. do you use the same approach if you have a pallet with 5x5x5 boxes stacked or if you have a pallet with 10x6x15 boxes stacked?)
- Which methods do you use for bulk items (Juran describes the use of a gridded board with small parts strewn onto the board)?
- Are there any standard SW solutions available?
Paul
Last edited: