Quantum XL’s AutoTest greatly reduces the time required to find and run the correct Hypothesis Test. However, there are some important notes that you should be aware of when using these tests.
Note 1: During the analysis, Quantum XL will test for normality using the Anderson Darling test. If the p-value is less than .05, then Quantum XL will assume the data is not Normal. In general, most statisticians would agree with this since we are greater than 95% confident that the data is not normal. However, if the p-value is greater than .05, Quantum XL will fail to reject normality and assume the data is normal. This is the primary area of concern. If the dataset is very small, the data may not be normal even though the p-value is greater than .05. If your dataset is small, you may want to use the more conservative approach and assume the data isn’t normal even though the AD test failed to reject normality.
The same is true for the test of equal variance. Quantum XL will run either the F-Test, Bartlett’s Test, or Levene’s test. If the p-value is greater than .05, then Quantum XL will assume the variances are equal.
Note 2: Quantum XL uses the Box Plot to determine the presence of outliers. The Box plot doesn’t provide a p-value for the presence of outliers, it either identifies them or fails to.
Note 3: You can download our flow chart that helps you choose the correct Hypothesis test and use it instead of the AutoTest.