I am doing a GR&R study with 10 parts, 2 operators, 3 replicates. After I process my GR&R data through MiniTab, the R-chart shows part 1 being out of control for operator A. Operator B is just within control for part 1. When I review my part*operator interaction chart, both operators lie on top of one another, except for part 1. I suspect the data for part one is an outlier and I want to replace part 1 with a different part. My quandry is not having any rules or guidelines for repeating/removing a part from a GR&R study.