STILL Mean to the Median...

[Leigh Meyers <leighcmeyers@xxxxxxxxx>]

>  U.S. median household income: $44,473 (three-year average, 2002-2004)
> (only 19 of 50 states are above this, by the way) (source: 
> http://www.census.gov)
.
Snipped from: 25 numbers journalist should know, an informal survey done 
on  [carr-l].

25 Numbers Journalists Should Know
by JTJ on Sun 13 Aug 2006 01:51 AM MDT
25 Numbers Journalists Should Know

A few days ago, I asked friends and colleagues on listservs to suggest 
25 relatively generic numbers journalists should know in order to be 
responsible, effective reporters and editors.  You sent along the great 
suggestions included below.  A handful of folks, however, responded to 
make two points:

   It is more important to know where to find pertinent numbers than it 
is to know specific numbers, and

   It is more important to know appropriate calculations â say, how to 
compute percent of change â that can be applied to specific numbers once 
they are found.

 Yes, points well taken.  But I donât think any of these are mutually 
exclusive.  Hereâs why.

 Any statistical analysis begins with classifying and counting.  That 
process is only relevant if put in some context.  If I tell you that 
Santa Fe, New Mexico has about 68,000 people, that number by itself has 
little meaning in terms of scale.  Is 68,000 big or small?  How do I 
tease some information out of that lonesome statistic?  Ah, but when we 
can ask how does it compare to other cities in the state, region or 
nation meaning and information start to bubble up? 

 The second analytic step is estimation.  This is helpful â perhaps 
necessary - to have some ballpark figure to help the analyst determine 
if his/her calculations are correct or âmake sense.â  If the city 
manager tells a reporter that the town has been growing by about .5 
percent per year since 2000, she could not estimate the amount of growth 
or its current aggregate unless she had a baseline number of 68,000.

 So we think that (a) journalists should always have some relevant â 
and fairly accurate â ballpark figures in mind to help with context 
(Yes, some of these will vary from beat to beat); (b) journalists should 
know where and how to find the historic and current statistics; (c) 
journalists should know how to do some fairly elementary arithmetic to 
tease information out of the data.

 Thanks to all for your contributions.

âTom Johnson [12 August 2006]

More: 
http://analyticjournalism.blogharbor.com/blog/_archives/2006/8/13/2226046.html