0
I am running an analysis where I need to find a reference value for small groups of observations (between 3 and 10). As some samples present at least a very discrepant value I will try to identify outlier. I found some references to modified z-score and also modified IQR however, the formulas differ and I have doubts about which formula should be implemented. Anyone have any idea? Is there a package / function that is more suitable for identifying extreme values in small-sized samples?
Your question is about statistics more than programming. A better place to seek an answer is the Crossvalidated (in English).
– Carlos Eduardo Lagosta