The 7 rules of Statistical Process Control (SPC), primarily used for interpreting control charts, serve as criteria to detect non-random patterns and identify when a process may be out of control or experiencing special cause variation. These rules help distinguish between common cause variation (inherent to the process) and special cause variation (attributable to specific, identifiable factors), prompting investigation and corrective action.
Here are the 7 common rules for interpreting control charts in SPC:
The 7 Rules of SPC for Control Chart Interpretation
These rules are often referred to as a subset of the Western Electric Rules or Nelson Rules and are crucial for effective process monitoring.
| Rule Number | Rule Name | Pattern Indicating Special Cause Variation |
|---|---|---|
| 1 | Point Outside Control Limits | One or more points fall outside the 3-sigma control limits (Upper Control Limit - UCL or Lower Control Limit - LCL). |
| 2 | Zone A Violation (2 of 3) | Two out of three consecutive points fall outside the 2-sigma warning limits (Zone A) on the same side of the centerline. |
| 3 | Zone B Violation (4 of 5) | Four out of five consecutive points fall outside the 1-sigma warning limits (Zone B) on the same side of the centerline. |
| 4 | Zone C (Consecutive Points) | 7 or more consecutive points on one side of the average (in Zone C or beyond). |
| 5 | Trend (Consecutive Trending) | 7 consecutive points trending up or trending down. |
| 6 | Mixture (No Points in Zone C) | 8 consecutive points with no points in Zone C. |
| 7 | Stratification (Many in Zone C) | 15 consecutive points in Zone C. |
Detailed Explanation of Each Rule:
Understanding each rule helps in diagnosing potential process issues.
-
Rule 1: Point Outside Control Limits
- Description: This is the most fundamental rule. If any data point extends beyond the calculated upper or lower control limits, it indicates a significant shift or disturbance in the process.
- Implication: This often signals a strong special cause affecting the process, requiring immediate investigation to identify and eliminate the root cause.
- Example: A sudden spike in defect rates going above the UCL.
-
Rule 2: Zone A Violation (Two of Three Points)
- Description: If two out of three consecutive data points fall beyond the 2-sigma warning limits (but still within the 3-sigma control limits) on the same side of the centerline.
- Implication: Suggests a process that is less stable, potentially starting to drift or being influenced by a new factor that is not yet strong enough to push a point beyond the main control limits.
- Example: Two out of three recent batches show measurements just above the +2-sigma line.
-
Rule 3: Zone B Violation (Four of Five Points)
- Description: If four out of five consecutive data points fall beyond the 1-sigma warning limits (but still within the 3-sigma control limits) on the same side of the centerline.
- Implication: Similar to Rule 2, this pattern suggests a creeping shift or a process that is becoming more biased towards one side of the average.
- Example: Four out of the last five products measured fall in the +1-sigma to +2-sigma range.
-
Rule 4: Zone C Violation (7 or more consecutive points on one side of the average)
- Description: This rule identifies a sustained shift in the process average. When seven or more consecutive data points lie entirely on one side of the centerline (either all above or all below).
- Implication: A prolonged run of points on one side of the average indicates that the process mean has shifted. This could be due to a new operator, a machine recalibration, or a change in raw materials.
- Example: Seven straight daily production counts are below the historical average.
-
Rule 5: Trend (7 consecutive points trending up or trending down)
- Description: When seven consecutive data points are consistently increasing or consistently decreasing.
- Implication: A trend suggests a gradual but persistent change in the process. This could be due to tool wear, equipment degradation, or accumulating impurities over time.
- Example: For seven production runs, the temperature readings gradually increased.
-
Rule 6: Mixture (8 consecutive points with no points in Zone C)
- Description: This rule is met when eight consecutive points are observed with none of them falling in Zone C (the area within one standard deviation of the centerline). This means all 8 points are outside the 1-sigma range.
- Implication: A mixture pattern often indicates that two different processes or conditions are operating alternately. This can lead to increased variability and make the process difficult to control.
- Example: Production output alternates between high and low values, with no values near the average, possibly due to two different shifts or material batches.
-
Rule 7: Stratification (15 consecutive points in Zone C)
- Description: When fifteen consecutive data points fall entirely within Zone C (the area within one standard deviation of the centerline).
- Implication: While seeming "good" (as points are close to average), stratification indicates an artificially reduced variation. This often occurs due to issues like incorrect data collection (e.g., averaging data before plotting), a faulty measurement system that isn't sensitive enough, or improper subgrouping. It masks true process variation.
- Example: All 15 recent product weights are almost identical, suggesting the scale might not be sensitive enough to capture real variation.
These rules provide a robust framework for identifying when a process needs attention, ensuring that improvements are targeted at actual special causes of variation rather than common cause noise.
