3 ms·
And a rather general observation - we expect "most extreme" anything to often be the result of measurement error or some sort of lie. If the variance of a disr
by roenxi 5y ago
And a rather general observation - we expect "most extreme" anything to often be the result of measurement error or some sort of lie.
If the variance of a disruptive event is higher than the variance of a process, we expect the extremes to be disruptive events. The things people are willing to say (by accident or as a lie) are much more radical than the results most processes work to. So it is not that strange if an extreme event turns out to be effectively a bad evidence keeping.