The masking of outliers problem was briefly mentioned in an answer here:
Outlier detection for heavy-tailed data.
These two requests are to help understand outlier masking better:
- Real-world examples of outlier masking, preferably multivariate where presumably masking is more prevalent.
- What outlier detection methods are resistant versus susceptible
to outlier masking, with accompanying explanation.
As an example of current understanding, it is easy to understand why Grubbs’s test uses an iterative approach and this would mitigate against masking.