Lee Sieswerda
> Transformed scales are used in the epidemiology literature,
> of course, but
> more often in the very technical literature. Indeed, I'm sure every
> conceivable scale transformation has been used at some
> point. However,
> epidemiology also serves wider, less mathematically
> inclined, public health
> and medical audiences. Most health care practitioners
> wouldn't know what to
> do with a log or square root scale. Also, epidemiological
> studies often end
> up in the media. Can you see an epidemiologist trying to
> explain the square
> root rate of influenza while on television? It is also very
> unlikely that
> newspapers will print charts with log or square root
> scales. So, if we are
> able to simply break the axis to make the scale more
> understandable to
> non-statistically-oriented types, then we do (and should).
No disagreement that there are issues when dealing with
lay audiences and/or the media.
I make two specific comments:
1. In many educational systems, square roots and logarithms
are, or were, taught at early ages. I can see that in
practice people don't retain what they don't use
but pandering to those with the weakest mathematical skills
is a very unsatisfactory way to present statistical science.
(I speak as someone with _no_ formal qualifications in mathematics
or statistics beyond high school.) That's probably
preaching to the converted.
2. Some serious journalism makes extensive use of log
scales. The Economist is one example, and indeed I have
found its use of graphics to be generally excellent,
and on the whole at a better level than many academic
journals.
Nick
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
