diff --git a/docs/FAQ/faq.md b/docs/FAQ/faq.md index 77cbc22da321ead5f6371a849381489934c272f8..aef036b60dae3db7bcd39ebd9565a1ba6e287480 100644 --- a/docs/FAQ/faq.md +++ b/docs/FAQ/faq.md @@ -75,6 +75,12 @@ You can find the contact information on the first page of this website. The KS test is generally used to compare continuous distributions, while the AD test can be used to test for a variety of distributions, including discrete distributions. Hence the AD test is used as the default test. The two tests are implemented as described in [Kolmogorov-Smirnov test](https://root.cern.ch/doc/master/classTH1.html#aeadcf087afe6ba203bcde124cfabbee4) and [Anderson-Darling test](https://root.cern.ch/doc/master/classTH1.html#aa395c473ea9693359a74189fbe0ee0db). + The probability you see in the plot is the so called `p-value`. + They are a statistical measure used to determine the probability of obtaining a result as extreme as, or more extreme than, the observed result if the null hypothesis is true. In this particular case the null hypothesis assumes that the two distributions (i.e. train versus test) are originating from the same underlying probability distribution (colloquially speaking are identical). + + In other words, p-values help to assess the strength of evidence against the null hypothesis. A low p-value suggests that the observed result is statistically significant, which means it is unlikely to have occurred by chance alone, and we can reject the null hypothesis in favor of the alternative hypothesis at a `confidence limit` given by `1-p`. + + However, it is important to note that p-values only provide evidence against the null hypothesis and not evidence in favor of the alternative hypothesis. ## Injection questions