The Xbar-R chart combines a mean chart (Xbar) and a range chart (R): the Xbar chart monitors the mean of each subgroup and reflects shifts in process location, while the R chart monitors the subgroup range (maximum minus minimum) and reflects shifts in process spread. It is the most widely used variable control chart in industrial production.
Use it when the data are variable (continuous measurements) and each subgroup has 2-8 samples, for example measuring 5 parts per shift or weighing 3 pieces per hour. With more than 8 samples per subgroup, switch to Xbar-s because the standard deviation uses more information than the range. With one value per subgroup, use I-MR.
Read the R chart first: an out-of-control range means within-subgroup variation has grown, so investigate measurement or process-dispersion causes first. Once the R chart is stable, read the Xbar chart: an out-of-control mean means the process center has shifted or a systematic factor exists. The tool flags points automatically by the eight run rules and gives a judgement hint.
Xbar chart: UCL/LCL = Xbar +/- A2*Rbar. R chart: UCL = D4*Rbar and LCL = D3*Rbar (for n less than 7, D3 = 0, so the R chart has no lower limit). A2, D3 and D4 are constants that depend on subgroup size n; the tool holds the constant tables and calculates automatically.