If a control chart displays a point outside the control limits, what is the appropriate action?

If a control chart displays a point outside the control limits, what is the appropriate action?

• A. Ignore the anomaly and continue testing.
• B. Recalibrate the instrument and re-run immediately.
• C. Recalculate the data and adjust the control limits.
• D. Investigate the cause and re-run the measurement if needed.

Answer: (D)

When a data point appears outside the control limits, it signals special-cause variation rather than normal random fluctuation. This means something in the process or the measurement system changed and needs to be identified before proceeding. The appropriate action is to investigate the potential cause—check instrument calibration, reagent lots, operator technique, environmental conditions, and sample handling—and then re-measure to confirm whether the process is back in control. Recalibrating and re-running immediately is not the default move because it assumes calibration issues without evidence and can mask real problems. Recalculating data and adjusting the control limits would undermine the chart’s integrity and hide genuine shifts. Ignoring the signal is unsafe because it could let a meaningful problem go unresolved. So, investigate the cause and re-run the measurement if needed.

When a data point appears outside the control limits, it signals special-cause variation rather than normal random fluctuation. This means something in the process or the measurement system changed and needs to be identified before proceeding. The appropriate action is to investigate the potential cause—check instrument calibration, reagent lots, operator technique, environmental conditions, and sample handling—and then re-measure to confirm whether the process is back in control. Recalibrating and re-running immediately is not the default move because it assumes calibration issues without evidence and can mask real problems. Recalculating data and adjusting the control limits would undermine the chart’s integrity and hide genuine shifts. Ignoring the signal is unsafe because it could let a meaningful problem go unresolved. So, investigate the cause and re-run the measurement if needed.