Control Limits
On a control chart, the upper and lower boundaries placed three standard deviations above and below the process mean (centerline) for a normally distributed measure. They show the range of variation expected from common causes and are not the same as specification limits.
Key Points
- Set at plus or minus three standard deviations around the centerline of the control chart.
- Indicate the expected common-cause variation of a stable, normally distributed process.
- Points beyond these limits signal potential special causes and a process out of control.
- Different from specification limits, which reflect requirements or tolerances, not process behavior.
Example
A project team tracks weekly defect counts. The average is 12 defects with a standard deviation of 2. The control limits are set at 12 minus 6 (LCL = 6) and 12 plus 6 (UCL = 18). A week with 19 defects falls outside the UCL, prompting investigation into a special cause such as a recent code change.
PMP Example Question
Which statement best describes control limits on a control chart?
- They are the customer-defined tolerance boundaries for acceptable results.
- They are fixed limits at two standard deviations from the mean for any process.
- They are statistically derived bounds, typically at plus or minus three standard deviations from the mean, indicating expected process variation.
- They are thresholds set by the project manager to meet schedule objectives.
Correct Answer: C — Statistically derived bounds at about plus or minus three sigma around the mean
Explanation: Control limits are based on process data and show the expected common-cause variation. They are not customer tolerances (specification limits) or arbitrary manager-set thresholds.
