We’ve all heard it before. Someone says, “If you want my opinion...” and they give it to you even after you’ve told them you don’t want it.
There are a number of reasons why you might decline the offer of an opinion. One may be because you consider that it won’t be worth the time to hear it. Or it could be that you don’t have the time to provide enough background so the opinion will have some value. You might not want an opinion because the person giving it knows nothing about the situation at hand. Did that ever stop the most opinionated from expressing one anyway?
When it comes to opinions on calibration reports there are two schools of thought. Some folks insist that calibration reports show whether an item passes or fails compared to some standard or specification. Others insist that we do not put such comments on their reports. There are good reasons for either position.
Those not wanting an outside calibration lab making decisions on their behalf may have been cautioned by their lawyer about the legal risks this poses. The last thing a lawyer acting on your company’s behalf in a product liability matter will want is a report with a negative opinion surfacing in the middle of it all. Especially when that opinion is irrelevant.
Some companies don’t want opinions on their reports because they may have developed their own criteria based on the tolerances they work to. If they get the data from the lab and know what they require there’s no need for a lab to give an opinion especially when the lab uses different criteria.
Companies that want ‘pass/fail’ decisions from their calibration source usually tend to use manufacturer’s specifications or published standards as their criteria for the decision. This is especially so when it comes to fixed limit plug and ring gages. Many do so without having read a copy of those standards and are unaware of details that could cause an item to be rejected even though it meets its functional requirements satisfactorily.
Published specifications are created for new equipment and are really the tolerances the maker of that equipment has to meet. People often cite them as their acceptance criteria for used equipment but this can result in perfectly useable equipment being scrapped unnecessarily. In the case of measuring instruments they can meet those specifications for some time after their initial purchase. However, when fixed limit gages are involved, a gage could be on bottom limit when new but wear outside of that ‘new’ limit with very little use but that doesn’t always mean it is no longer useful.
Opinions based on acceptance criteria that are not determined by what is actually required from a technical point of view will be viewed as gospel when they are given by a calibration source as far as too many quality auditors are concerned. It’s the easy way out to deal with the subject but it is usually a bad choice—just like opinions based on them.
Another bomb with a short fuse is how the laboratory treats measurement uncertainty when giving an opinion. There are several ways to deal with it, but if you don’t specify which is to be used, they will ignore it although it should be noted on the report. Like any measurement, if the uncertainty is not considered any opinion will be of little value. Even when uncertainty is taken into account it can put a significant amount of acceptance criteria into doubt. This is regularly encountered when arbitrary numbers have been chosen as acceptance limits and go head to head with limitations on the state of the art.
To make matters worse, acceptance limits borrowed from new product specifications are sometimes not provable but have been accepted as ‘the standard’ by one industry or another for years.
This is problematic with some old standards where advances in dimensional metrology indicate they are not measurable with any degree of certainty.
I suspect some folks may be expecting an opinion on all of this and how to sort it out but since you didn’t ask for my opinion I’ll follow my father’s advice of long ago and keep my opinions to myself. At least for this column.