I am developing an in-lab audit program of calibrations completed. This is a hands-on audit including re-checking of selected parameters to verify by an objective, qualified calibration technician or engineer that steps were correctly checked. For example, a sample size of 5 calibrated items per week, from different technical areas of the lab, and/or done by different technicians, and a selection of parameters (not a full replication of the cal, a selected critical set of parameters, such as on a frequency counter, re-checking timebase and top end of frequency response on at least one input channel). Sample size would be periodically increased or decreased based on results. I haven't nailed down sample size or details. I have been digging to try and find some industry standards on this. Anyone here have any ideas? You may also email me at rn0414@email.sps.mot.com

I don't know of any 'industry standards' off hand. Maybe one of the other folks does. I would assume this is more of a statistical problem. You have x total calibations. You decide on the confidence level you want and from that you determine a sample size.