Test & Measuring Equipment - Statistical studies to calibration determine frequency

This question is a result of a survey by one of our customers.

The question:

Control of Test & Measuring Equipment.

"Requirement by our customer.
Calibration results to be analyzed using statistical techniques to establish frequency".

I know of no requirement for ISO or QS that has any language concerning statistical studies for calibrations to determine frequency.

Would appreciate any help. Thanks

[David Mullins]

Good Management!
When you open the book that came with the equipment, what is the manufacturers recommendation on calibration frequency. Determining this frequency yourself can be an expensive and time consuming operation. Additionally, I know in Oz there are already numerous standards that provide this sort of guidance.
At the end of the day, the stats to determine frequency relate to the average time taken to exceed the acceptable degree of risk of the accuracy of the equipment.

------------------

[Marc]

I don't know that there is a specific requirement for statistical techniques but I would be ready to use cal records as evidence of stability during your cal interval. I would also be ready to say you evaluated the need for statistical analysis of calibration data and found it to not be value added based upon instrument stability. You don't need a graph to tell when the instrument is wearing or such - its stability will fall with use. If heavily used this could be a day or a week. Or it could be a month or a year. Lot's of possibilities.

[Marc]

By the way, graphing, as I understand it, is a 'statistical technique'. Have any of you run into this?

Because of it being calibration in nature, I'm moving this thread to the Calibration forum.

This link will take you there: Statistical studies for calibrations to determine calibration frequency?

[This message has been edited by Marc Smith (edited 04 February 2000).]