Wheeler

Wheeler’s XmR Chart (Individuals and Moving Range)

Wheeler argues that most business data (monthly sales, daily output, wait times) comes as individual values.

The XmR chart consists of two linked plots:

X Chart: Plots the individual data points.

mR Chart: Plots the "moving range" (the absolute difference between consecutive points).

This method is robust because the moving range provides a logical estimate of the process variation even when you have a small number of data points.

Key Principles from Wheeler's "Understanding Variation"

The Fallacy of Comparisons: Comparing "This Month vs. Last Month" or "This Year vs. Last Year" is often meaningless. Unless the difference exceeds the control limits, it is just noise.

Numerical Goals: Setting targets without understanding the process capability is what Wheeler calls "Management by Fear." If a process is in control but the average is below a target, asking for "more effort" won't work—the system itself must be changed.

The Four Possibilities:

Ideal State: Process is in control (predictable) and meeting specs.

Brink of Chaos: Process is in control but not meeting specs.

Threshold of Chaos: Process is out of control but meeting specs (for now).

State of Chaos: Process is out of control and not meeting specs.

Interpreting the Signals

According to Wheeler and the Western Electric rules, a process is "sending a signal" (showing special cause variation) if:

A single point falls outside the control limits.

At least 3 out of 4 consecutive points are closer to a limit than to the average.

Eight consecutive points fall on the same side of the average (a shift in the mean).

By following Wheeler’s methodology, organizations stop "chasing ghosts" (reacting to noise) and focus their energy on investigating genuine signals or fundamentally improving stable processes.