Thom <[EMAIL PROTECTED]> wrote in news:[EMAIL PROTECTED]:

> Some of us remember you! I think your point about avergaes is well made.
> I'm sometimes amazed at how people struggle with percentages - until I
> remind my self that it is a _tricky_ concept. Percentages and other
> averages seem to be difficult because the process of arriving at the
> answer strips away the context. So a 29% increase could mean x 1.29 or
> an increase from 10% to 39%.

And even when talking about percentages as proportions, some people seem to 
forget that the numerator has to be the count of a subset of the 
denominator.  I've seen too many people try to estimate the prevalence of 
AIDS in <group> by taking the total number of members of <group> who have 
been diagnosed with AIDS since 1981 and dividing it by an estimate of the 
number of <group> members in the current population.  This gives wild 
overestimates because it doesn't take into account the fact that a majority 
of the people who have been diagnosed with AIDS since 1981 are ex-people 
and hence aren't counted in the denominator.
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to