A Baseline Measurement is needed to determine the exact starting point for each project contract primary (Y) and secondary metrics (y's). These are usually referred to as the "Before" measurement. The final project measurement is called the "After" measurement.
The primary metric of interest (to improve) is the output, Y. The baseline measurement (or beginning performance) is necessary to determine the progress of the team. Going through his process from the most recent data collection and sampling may generate a value different from that shown originally on the project contract.
At this point, if the difference is significant, the Champion, Sponsor, and management may need to revise the contract and the financials.
A characteristic, opportunity, or feature that is non-conforming. Each unit or piece being appraised may have multiple defects.
To distinguish the spectrum of defect types among units write clear instructions, use visual aids, and/or add numerical specifications.
Definitions of defects should be known and held consistent throughout the project. If it is found that new instructions and aids are necessary then these should be shared (such as in a SOP or Work Instruction) with all operators and employees affected so appraisals are reproducible among everyone.
A Poisson comparison is used for before and after metrics when counting defects per unit since there may be more than one defect per unit that creates a defective unit.
A piece or unit that is non-conforming or does not meet the customer specification. A defective part or unit may have one or more defects.
A part can be either one of two outcomes: Defective or Not Defective
When there are two outcomes and equal chance each time, this follows the Binomial Distribution
If there are 80 critical characteristics on a part, then one part can have up to 80 defects. Whether the part has 1 or 80 defects, it is considered one defective part.
Numerical statistics are used as baseline measurements, examples are shown below and calculations can be found at Basic Statistics
DPMO is not the same as PPM since it is possible that each unit
(part) being appraised may have multiple defects of the same
type or may have multiple types of defects. A part is defective if it
has one or more defects. The number of Defectives can never exceed the number of defects.
If each part only has one characteristic that can be a defect, then DPMO and PPM will be the same.
DPMO will always exceed or equal PPM for a given yield (sigma level) performance.
Assessing continuous data is preferred over attribute data, when possible convert to variable data. Follow the flowchart shown below to determine process capability when analyzing continuous data.
Six Sigma Modules
Green Belt Program (1,000+ Slides)
Cause & Effect Matrix
Central Limit Theorem
1-Way Anova Test
Correlation and Regression