A tool for averaging the capabilities of a family of related nonsequential steps in a process or a family of related components. For example, you may want to determine the average capability for processing loan applications across several different bank branches that use essentially the same process. Or, you may want to average the capabilities for different color door handles produced by the same process. The steps or components to be averaged can have either discrete or continuous data, although one average is reported for all discrete outputs and a separate average for all continuous outputs. The report also shows detailed capability metrics for each step or component (for example, step throughput yield and DPMO).

Answers the questions:

- What is the average capability (DPMO) of a family of related nonsequential process steps?
- What is the average capability (DPMO) of a family of product components?
- Which process steps, subprocesses, or subassemblies have the greatest need for improvement?
- What is the average capability of the process (both long-term and short-term) at the start of the process improvement project?
- What is the average capability of the process (both long-term and short-term) after you have made improvements?
- What is the capability for a single process step with discrete data?
- What is the probability of performing a step or producing a component defect-free the first time?
- What is the probability of performing all steps or producing all components for a product or service defect-free the first time?

When to Use | Purpose |
---|---|

Pre-project | Assist in project selection by focusing on portions of a process or a product that have high defect rates. |

Start of project | For processes with discrete data, perform a baseline capability analysis on the process to determine the average capability (DPMO) of the overall process and the capability of the individual steps or components at the start of the project. The baseline analysis also helps set improvement goals for the project. |

Start of project | For processes with continuous data, average the capability metrics (DPMO) across a family of related components to obtain a baseline average DPMO for the family. |

Mid-project | For processes with discrete data, perform a confirmation capability analysis after improvements have been implemented to confirm that the individual process steps and components, as well as the overall process, performs as predicted prior to implementing controls. |

Mid-project | For processes with continuous data, average the capability metrics (DPMO) across a family of related components after improvements have been implemented to confirm that the average capability (DPMO) of the family meets expectations prior to implementing controls. |

End of project | For processes with discrete data, perform a capability analysis after implementing controls to obtain a final assessment of process capability, and also determine whether the improvement goals of the project were attained. |

End of project | For processes with continuous data, average the capability metrics (DPMO) across a family of related components after implementing controls to confirm that the average capability (DPMO) of the family meets expectations. |

Summarized data: defects, opportunities, units, DPMO, and Z shift.

- The report has two tables: one for discrete outputs and the other for continuous outputs.
- When you are averaging the capabilities of related process steps of product components, all of the steps or components should have the same type of outputs - either they are all discrete or they are all continuous.
- Enter the name of each step or component into the appropriate table.
- In the Discrete Outputs table, you must enter (for each step or component) the number of units tested, the number of defects observed, and the number of opportunities for defects per unit.
- In the Continuous Outputs table, you must enter (for each step or component) the DPMO (long-term), which you usually obtained from a capability analysis, such as the Capability Analysis (Normal) analysis capture tool in Quality Companion.
- For both discrete and continuous outputs, you may also enter a value of Z shift, which is the difference between Zbench (ST) and Zbench (LT). This value is optional; however, if you delete the default (1.5), you will not obtain a Zbench (ST). If you do not know the value of Z shift, which is often the case, many Six Sigma deployments use a value of 1.5, which is the default used in both tables.

- The Capability Averaging Report requires that you have accurate counts of defects, units, and the number of opportunities per unit for any step or component with discrete data. The opportunities per unit can be the most difficult to determine; the lower the opportunity count, the more critical it is to be accurate.
- The report assumes that the proportionate number of units tested for each process step of product component is approximately equal to the proportions that you would observe in the day-to-day operation of the process,
- The Capability Averaging Report does not make use of data recorded over time to show process stability. Capability Sixpack (normal or nonnormal), Capability Analysis (Binomial), and Capability Analysis (Poisson) use process data over time. The major difference between them is, using the Capability Averaging Report, you can enter data for multiple process steps (either continuous or discrete) and combine those data for multiple steps into a single overall capability analysis, while the other reports depict capability for a single step over time.
- The step throughput yield (YTP) column in the Capability Averaging Report calculates the throughput yield for each step/component, which is the probability of correctly performing the step or building the component with no defects the first time (no scrap or rework). YTP is the opportunity-level DPMO raised to the power of the number of opportunities per unit.
- Below the Discrete Outputs and Continuous Outputs tables, the report displays a Combined Metrics table for that set of outputs, which includes the following metrics:
- Total defects, which is the sum of all defects for the steps/components included in the set of outputs.
- Total opportunities, which is the sum of all opportunities for the steps/components included in the set of outputs.
- Average DPMO, which is the total defects divided by the total opportunities, then multiplied by 1,000,000. The average DPMO is the defect rate, at the opportunity level, for either the set of discrete outputs or for the set of continuous outputs, depending on which table was used.