From: mat green
Date: 11/15/2005
Time: 5:40:55 AM
hi there, i wonder if anyone could help? i am working for a label printing company, whose quality needs to improve. currently during a production run, 3 samples are taken during a run - at the start, middle and end. these are random samples (the label may be printed up to 10 across, and the operator usually removes one label from the edge of the reel as this is the easiest position to remove a label from the moving web). on the basis of those 3 samples, which may come from a production run of 1000 labels, or may be 10,000 depending on the order. how can we improve the sampling regime so that we have a higher confidence that the samples we are taking actually represent the quality of the whole production run - we are currently rejecting a lot of defective labels. if anyone can help explain statistical sampling in "laymens terms" to me, it would be a great help. email me at gigtickets@gmail.com