Why can averaging many Cauchy observations fail to stabilize around a finite population mean?
StatisticsLevel: IntermediateContinuous probabilistic models
Why can averaging many Cauchy observations fail to stabilize around a finite population mean?
Hints
- Check whether the mean exists.
- Heavy tails can invalidate familiar averaging arguments.
Answer
The Cauchy distribution has no finite mean or variance. The ordinary finite-moment law-of-large-numbers reasoning therefore does not apply.
Source and licence: SciPy — continuous probability distributions · BSD-3-Clause · answer: Quantoros editorial