Re: Type 1 Gauge R&R Excel Template anyone?
You definitely want only one appraiser, but as I mentioned before, using 1 part only looks at a very narrow part of the gage. For example, lets assume as one possible scenario, that your measurement system is slightly "chunky." This is a normal condition of variable data caused by the discrete scales we impose on continuous functions (i.e., our time readings are usually in seconds, but if we want, we can measure to nanoseconds or some other fraction, but sooner or later we settle at some increment. When we measure a single part, it is going to have a true value that is probably slightly larger or smaller than our nearest increment. If it is very close to an interval, it will give us a bunch of identical readings. A little further away and we may get 2/3 of the values as one reading, and the other third as the second closest increment. This is one example, but the same kind of problem holds true with other aspects of measuring a single part.
When we look at multiple parts, still only with a single appraiser, but representing a broader pool of possible part situations, we get a type 1 gage error that is less affected by idiosyncracies of the product.
As far as a P value goes, this does not exist in isolation. If you measure 1 part 50 times, you will have within part variation. What are you going to compare this to? Any value only has meaning in its context within other numbers.
