Degenerate point of dispersion estimators

1 · Andrey Akinshin · July 12, 2022, 9:24 a.m.
Recently, I have been working on searching for a robust statistical dispersion estimator that doesn’t become zero on samples with a huge number of tied values. I have already created a few of such estimators like the middle non-zero quantile absolute deviation (part 1, part 2) and the untied quantile absolute deviation. Having several options to compare, we need a proper metric that allows us to perform such a comparison. Similar to the breakdown point (that is used to describe estimator robustn...