J
Alex, I quickly scanned your Excel sheet, there doesn't seem to be anything wrong with this.
As regards %Tol, IMO, there should not be a very BIG difference between 2nd and 3rd editions.
The 2nd edition directly calculated the value of the "Full spread" of GRR error, which was 5.15* GRR_Standard_Deviation. %Tolerance = Tolerance / (Full GRR) = (Tol)/(5.15*GRR_Stdev) = (Tol/5.15) / (GRR_Stdev)
In the third edition we calculate Standard_deviation of GRR (and also EV,AV, TV,PV etc). % Tol = Tolerance / (Full GRR), where full GRR spread is 6*GRR_StDev. This is the same as (Tol)/(6*GRR_Stdev) = (Tol/6) / (GRR_Stdev)
The difference in %Tol will only be to the extent of using 6 in place of 5.15.
As regards %Tol, IMO, there should not be a very BIG difference between 2nd and 3rd editions.
The 2nd edition directly calculated the value of the "Full spread" of GRR error, which was 5.15* GRR_Standard_Deviation. %Tolerance = Tolerance / (Full GRR) = (Tol)/(5.15*GRR_Stdev) = (Tol/5.15) / (GRR_Stdev)
In the third edition we calculate Standard_deviation of GRR (and also EV,AV, TV,PV etc). % Tol = Tolerance / (Full GRR), where full GRR spread is 6*GRR_StDev. This is the same as (Tol)/(6*GRR_Stdev) = (Tol/6) / (GRR_Stdev)
The difference in %Tol will only be to the extent of using 6 in place of 5.15.
do we use the 5.15 even if we are conducting a GR&R study when the %R&R is calculated at 95% confidence?