In This Topic
Step 1: Examine the MR chart to determine whether the process variation is in control
Before you interpret the Acceptance I chart, examine the Moving Range chart (MR chart) to determine whether the process variation is in control. If the MR chart is not in control, then the estimate of process variation and the acceptance control limits on the Acceptance I chart might not be reliable.
The MR chart plots the moving ranges. The center line is the average of all moving ranges. The control limits on the MR chart, which are set at a distance of 3 standard deviations above and below the center line, show the amount of variation that is expected in the moving ranges.
Red points indicate moving ranges that fail at least one of the tests for special causes and are not in control. If the same point fails multiple tests, then the point is labeled with the lowest test number to avoid cluttering the graph. If the chart shows out-of-control points, investigate those points.
Out-of-control points can influence the estimates of process parameters and prevent control limits from truly representing your process. If out-of-control points are due to special causes, then consider omitting these points from the calculations. For more information, go to Specify how to estimate the parameters for Acceptance I-MR Chart.
The MR chart does not show evidence of unusual variation, so the engineer can interpret the Acceptance I chart.
Step 2: Examine the Acceptance I chart to determine whether shifts in the process mean are acceptable
The Acceptance I chart plots individual observations. The center line is an estimate of the process average. The acceptance control limits are based on the specification limits, the accepted percentage outside each specification limit, the probability of intervention, and the estimated process variation. The chart displays a lower acceptance control limit, an upper acceptance control limit, or both, depending on which specification limits you enter.
Red points indicate observations that fail Test 1 because they are outside an acceptance control limit. A point outside an acceptance control limit indicates that the process mean shifted far enough toward the corresponding specification limit that the accepted percentage of process output outside the specification limit might be exceeded.
Shifts in the process mean that remain within the acceptance control limits do not indicate that the accepted percentage is exceeded. If the chart shows an out-of-control point, investigate whether an additional cause or an increased effect from an existing cause produced the shift.
Out-of-control points can influence the estimates of process parameters and prevent control limits from truly representing your process. If out-of-control points are due to special causes, then consider omitting these points from the calculations. For more information, go to Specify how to estimate the parameters for Acceptance I-MR Chart.
Two observations are above the upper acceptance control limit. These points indicate that the process shifted far enough toward the upper specification limit that more than 1% of the process output might exceed the specification limit.
After observation 100, the process mean shifts upward, but most observations remain within the upper acceptance control limit. The engineer should investigate the two out-of-control observations to determine whether the shift reflects an additional cause or a stronger effect from an existing cause. Possible causes include a machine adjustment, a change in raw material, an operator change, or a different production lot.
Step 3: Identify which points failed each test
Investigate any observations that fail a test. On the Acceptance I chart, Test 1 identifies observations that are outside the acceptance control limits. On the MR chart, Tests 1 through 4 detect special causes in the process variation.
If a point on the MR chart fails more than one test, the graph displays only one test number. The test results in the output identify all the tests that each point failed.
Test Results for I Chart of Fill weight deviation
| TEST 1. One point outside of acceptance control limits. |
|---|
| Test Failed at points: 112, 154 |
Note
When you use several tests on the MR chart at the same time, the sensitivity of the chart increases. However, the false alarm rate also increases, which can make you react to the test results unnecessarily.
For more information about the tests for special causes on the MR chart and when to use them, go to Using tests for special causes in control charts.
