[R] regression towards the mean, AS paper November 2007

hadley wickham h.wickham at gmail.com
Tue Dec 18 18:54:43 CET 2007

> Thanks for the example, Hadley.  To me, this suggests we should stop
> teaching histograms in Stat 101 and instead use quantile plots, which
> give excellent results for n=100 and even surprisingly good results
> for n=10:

It all depends on what you're trying to do - I don't think histograms
are particularly good as density estimators, but that's not what
you're using them for most of the time! You're using them as an
exploratory tool to try and understand what's going on in your data -
often you'll need to use very small bin widths which help find
unexpected gaps and patterns in your data.   It's helpful to have some
feel for what common distributions look like.



More information about the R-help mailing list