Select distribution parameters to test for Parametric Distribution Analysis

Stat > Reliability/Survival > Distribution Analysis > Parametric Distribution Analysis > Test

Use to test whether a distribution parameter for a sample differs from a specified value, or whether a distribution parameter differs for two or more samples. You can only use these options when you select Maximum likelihood from Estimation method on the Distribution Fitting sub-dialog box.

Test parameters

Test whether parameters equal a specified value. The values can be the parameters of the historical distribution, such as those that were determined from previous studies. Therefore, you may test the hypothesis that the new data comes from the same population (same distribution) as the data collected in previous studies.

For multiple samples, the test simultaneously evaluates whether all of the parameters equal the value that you enter.

When the data have multiple failure modes, enter a value for every parameter. For example, for data with 4 failure modes that all follow the Weibull distribution, enter a value for every shape parameter.

Test equality of parameters

You can test the equality of parameters across different samples when the analysis has multiple samples and has no failure modes.

To test whether two or more samples come from the same population, select the first two options for 2-parameter distributions, and select all three options for 3-parameter distributions.