For Gauge R&R to work it is important that our data is normal distibuted. If we don't have enough data the it might not be normal distributed. We can test if the data is normal distributed using a Shapiro-Wilk Test. Small values of W are evidence of departure from normality. It is important that our parts are normally distributed. A P-value smaller then 0.05 indicates that the data is not Gaussian.

For more information take a look at: