Now that we’ve explored using gummi bears to do a gage Linearity and Bias Study and a Type I Gage Study, it’s time to use gummi bears to practice the third and final type of measurement systems analysis that I’m planning to demonstrate: the gage R&R study.
"R&R" stands for Repeatability and Reproducibility, which are the two sources of variation we typically evaluate in a gage R&R study. Repeatability assesses how well the same person can get the same measurement over and over again when he or she measures the same part. Reproducibility assesses how well different people can get the same measurement for the same part.
But first, we’ll want to get a data sheet set up for recording our measurements, something that Minitab makes easy. We’ve already used Create Gage R&R Worksheet once to prepare for the linearity and bias study. We’ll use it in a more traditional way this time.
- Open Stat > Quality Tools > Gage Study > Create Gage R&R Study Worksheet.
- Choose the Number of parts that you’re going to measure. (I’m going to use 3 here, because I’m going to ask my coworkers to volunteer to be extra operators, and I think they’ll be more likely to play along if I don’t ask them to do too many measurements. In real life, we want to study the range of outputs we expect to see from the process.)
- Add descriptive text for the parts. (I used 12 inches, 30 inches and 48 inches.)
- Choose the Number of operators. (I’m going to use 4, even though I haven’t gotten any official volunteers yet. I’m an optimist by nature.)
- Add the names of the operators.
- Choose the Number of replicates. Click OK. (I’m going to choose 2 here, again mostly as a kindness to my volunteers.)
Now we have a worksheet setup that will let us keep track of our operators and the distances of the gummi bears they’re going to measure.
Of course, in your real measurement systems, you may have other sources of variation that you want to study besides just repeatability and reproducibility. If you’re ready for more, check out this article: Gage R&R Study (Expanded). But for us, it’s one thing at a time. We’ll get ready to look at the results from the gage R&R study next time.