MSA reaction plan if a test fails

If you 'fail' the MSA, you need to understand what failed and then what caused it to fail and then you would take corrective action and repeat the .
for example if you fail the Reproducibility portion you would determine which operators were the most different from each other and observe their technique. it is possible that one operator dos it really well and another is not so skilled. In this case you can train the lesser skilled operator to be like the more skilled operator. If you fail repeatability you can again look to see if there any operators that are worse or better than the others or if they all have poor repeatability. In this case again observation of technique is critical. Another contributing factor to repeatability is within part variation. if you are randomly measuring a characteristic without regard to the measurement location it is possible that the lack of repeatability is due to within piece variation. You can determine this by marking the part and always measure the same location in the MSA study. If this is the cause, then the corrective action may be to always measure the part at a set location, correcting the within piece variation, or changing the acceptance criteria to accommodate multiple measurements on the piece (among other options). It will depend on the use of the part...