Hi Leonard,
I am somewhat disturbed by the new 3D pie-charts (and the possibility
of further 3D charts).
What in particular disturbed you about the 3d charts in the new chart
(That's what I suppose you mean by new 3D pie-charts)? There are some
differences to the old chart, but no major visual difference. So,
basically, the 3d charts are the same as before. Maybe this shows that
our new wizard is better than the old one, because you find the
possibility to get 3d charts. In the old wizard you had to scroll the
first page to get the 3d charts, which was not very obvious.
Unfortunately, 3D charts DO NOT add any statistical benefits and
should be strongly discouraged!!! This view is held by many renowed
statisticians, including the late Prof. Alvin Feinstein. I have
*briefly* discussed this issue on
http://wiki.services.openoffice.org/wiki/Statistical_Data_Analysis_Tool
, and I hope to have more time to expand it even further (see also
Feinstein AR. Principles of Medical Statistics. Chapman & Hall 2002)
Of course it does not add information, it's just that people like 3d
charts for presentation. Imagine what would happen if we suddenly
dropped the support for all 3d chart types. We can simply not do this.
And besides, we don't want to.
The basic feature of 3D pies is to represent accurately the magnitude
of the variable. The *AREA* of the pie is directly proportional with
the variable of interest (and so it is proportional with the outer
length). In a 3D pie, everything is distorted (to create the
3D-effect), so the area, the outer length, and the perception of these
values is NO LONGER CORRECT!!!
That depends on your perception. If you view the 3d pie as a 2d image
and try to convince your brain that it is not 3d what you are looking
at, your are right. But if you have a real pie, a cake, and cut
differently sized pieces, would you say that a viewer is no longer be
able to interpret the sizes correctly, because it is 3d? Or if he would
be able to interpret them correctly, would this no longer be the case
when he would look at a photograph of the pie? The brain is a pretty
amazing tool (I am sure most children would find out if a cake cut into
12 "equally sized" pieces if there was a piece that is slightly bigger
than the others).
There are examples of a wrong usage of 3d. E.g. I remember a chart where
amounts of oil (the volume) were illustrated by differently sized
barrels. The different sizes were modeled as the height of a 3d picture
of a barrel, which scaled proportionally. As a result the volume of the
barrels changed much more than was appropriate. (Double the height means
double the volume only when the radius of the ground circle remains the
same). But this is a different story. I think we don't offer charts
where you have this kind of misusage of 3d.
Of course, you should always use 3d charts with care. If you have a 3d
pie which you view almost from above, there is no big difference to a 2d
pie, except that some people find this more appealing. If you view a 3d
pie almost from the side, you won't see much. It is up to the creator of
a chart to make sure that the viewer can see the relevant data. Of
course some people make use of such things to make the data look
"better", e.g. by putting the competitors large pie segment in the back,
so that it might look smaller. But people fake their charts anyway.
There are many ways to do so. Offering no 3d would not solve this problem.
There is NO NEED for a 3rd dimension in pie charts (it simply does NOT
bring further information). Therefore, the user should be WARNED, IF
selecting a 3D chart, that this is statistically wrong and unnecessary.
It is not true that 3d charts are statistically wrong. It is, however,
correct, as you stated before, that the perception of a user might be
wrong. I think this is a difference. If 3d charts would be wrong we
wouldn't offer them.
So, your point that 3d is problematic and not really necessary is
correct. But we won't be able to drop 3d chart support for two main
reasons: 1. compatibility and 2. people want them.
Regards,
Bjoern
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
