You have a substantiation test plan because you are testing whether or not a redesigned unit is an improvement over the existing system. Minitab uses the following planning values to create the test plan:
- Distribution: The distribution that appropriately models failure times for your system.
- Shape (or Scale): The Weibull shape parameter, or the scale parameter for other distributions.
- Scale Goal: The minimum value that you want to demonstrate.
- Actual Confidence Level: If you specify the testing time, the calculated sample size is rounded to an integer value, which makes the actual confidence level slightly higher than the target confidence level.
Substantiation Test Plan
Distribution: Weibull, Shape = 1.3
Scale Goal = 1000, Actual Confidence Level = 95%
Failure Sample Testing
Test Size Time
0 3 998.906
For the ball bearing data, you selected a Weibull distribution with a shape of 1.3. The value you want to demonstrate for the Weibull scale is 1000 hours. Because you provided a sample size and Minitab calculates a testing time, the actual confidence level is equal to the target confidence level of 95%.