[R] Michaelson-Morley Speed of Light Data

[Křištof Želechovski]

Dnia środa, 11 kwietnia 2012 13:26:43 Prof Brian Ripley pisze:
> On 11/04/2012 12:42, Duncan Murdoch wrote:
> > On 12-04-11 12:43 AM, Křištof Želechovski wrote:
> >> <URL:
> >> http://finzi.psych.upenn.edu/R/library/datasets/html/morley.html>
> >> 
> >> "The classical data of Michaelson and Morley on the speed of light"
> >> 
> >> Can you provide more information about the data? How were they
> >> obtained,
> >> etc.? I do not have the book "Genstat Primer" and the nearest location
> >> where
> >> it is available is University of York which is rather far from my
> >> location.
> > 
> > If you can't find the cited reference, I'd try Google. For instance, it
> > led me to this page
> > 
> > http://en.wikipedia.org/wiki/File:Michelsonmorley-boxplot.svg
> > 
> > which appears to show five series.
> 
> Yes, but that is derived from R.
> 
> AFAIR the history, Bill Venables got this from Weekes (1986), a book I
> have only ever seen in Adelaide.  A better reference is
> 
>       S. M. Stigler (1977) Do robust estimators work with real data?
>       Annals of Statistics 5, 1055–1098. (See Table 6.)

The data in R are identical with Table 6 but they were not obtained by 
Michaelson and Morley; they were obtained by Michelson.  The description is 
wrong.

It is also a problem in R itself, as evidenced by the following instruction:

data(morley)

Of course, you can name a data set whatever you like, even data(marilyn) would 
be all right, but may I suggest that this data set be renamed to michelson?

Please fix,
Chris