1. Home
  2. Glossary
  3. Control Chart

Control ChartShewhart Chart

A time-series chart with statistical control limits used to distinguish normal variation from special-cause variation.

  • Six Sigma
  • Quality

A control chart plots a process metric over time against a center line and upper/lower control limits derived from the data itself. It is the core tool of Statistical Process Control.

It answers the most important question in process management: is this variation normal "common cause" noise to be left alone, or a "special cause" signal that demands investigation? Reacting to common-cause noise (tampering) actually increases variation.

Example. Shaft diameters are checked in subgroups of 5 each hour. The grand average is 25.00 mm and the average range is 0.04 mm. For an X-bar chart with n = 5 the limits are X-bar ± A2 x R-bar = 25.00 ± 0.577 x 0.04, so 24.977 to 25.023 mm. These are set from the process, not the tolerance; a process can be in control and still make scrap if the limits sit outside the specification.

Reading it: a point beyond a limit, 7 points on one side of the centre line, or 6 points trending in one direction are signals worth stopping for. Adjusting the machine after every point that is merely above average adds variation. The chart type follows the data: X-bar and R for measured values in small subgroups, I-MR for single readings, p and c charts for counts of defectives and defects. See the SPC guide.

Go deeper

Related terms

Browse all glossary terms