"Just for the record", there is one apparently statistically sound use of
Excel's doughnuts -- see the article
Bordley, R.F., Representing trees using microsoft doughnut charts, The
American Statistician, 2002, vol. 56, no. 2, p. 139-147.
I just regret that the figures are rather sloppily drawn and that the macro
supposedly available for public download is nowhere to be found. (Anyone out
there able to provide it? I'd be really grateful.)
[Digression: As someone daily dealing with publishing in more or less "top"
scientific journala, I cannot help myself being left with a somewhat bitter
taste in the sense that it might very well have to do with the author being
a General Motors executive in addition to affiliated to two US universities,
while -- trust me from first-hand and second-hand experiance -- even such
generally trustworthy and impecaple journal as TAS exercises a whole lot
more severe technical and also content standards in other cases. But don't
get me wrong -- the article is OK and I like and support the idea.]
As for "getting information into people's heads", "people's judgment" and
the like, I would just like to add a note of caution as a semi-qualified
psychologist of perception (BSc & MSc) and statistician (PhD submitted). --
Sure, promoting Tufte's (which I fully and publicly and regularly do) and
Cleveland's teachings (erm, khm, problems with psychometric findings
commencing, though from the purely statistical point of view he undoubtedly
is another hero of data visualisation) is fine, but one should be aware of
the "human tendency" (I know, now I, a notorious Pompous European
Intellectual, am generalising and writing simplistic and <American> stuff)
towards the "pendulum paradigm" of swinging from one extreme to the oposite
before finding the balance ...
In short, as with just about anything, there's much more to pies. And if you
(I) need a true hero from this (my) field, it should be Leland Wilkinson and
his Grammar of Graphics. (Just for the record, he wrote about <pies> earlier
and it's even accessible from his website, but since there is a zillion of
other and wider issues to study here, the proverbial interested reader
should just start with the book).
Cordial regards,
Gaj Vidmar
Univ. of Ljubljana, Fac. of Medicine, Inst. of Biomedical Informatics
http://www.mf.uni-lj.si/ibmi-english [/biostat-center]
[-> Research] & [-> Software [-> Excel]]
