"Eric F. Richards" wrote in message
...
"Steve Stone" wrote:

I'm a database analyst by day and I know statistics can be made to say
anything you want them to say, especially if you ask the wrong questions
that reflect what the reviewer wants to hear and not what the public
wants
to tell them.

I tried making that point a couple months ago, with no affect.
Everyone thinks that any collection of data can be analyzed with a
normal distribution... and it just ain't so.

You and steve miss the point . Radio staitons have no reason to order bad
research. Jobs depend on increasing or holding ratings. Very good companies
are used, and they spend lots of time avoiding the pitfalls you mention.

Likewise, like you say, surveys are often -- perhaps usually --
slanted to return the results they want. My personal experience with
Arbitron left me unimpressed.

Advertisers have a committe that audits them. That is adequate for them to
spend about $21 billion on radio advertising. Advertisers seem to believe
the nature of Arbitron ratings far more than your rather distorted and
inaccurate to the Nth degree analysis of thier function and methodolgy (you
do not even get the terms of the trade right).

The whole radio ratings game is a self-serving, narrow minded exercise
in mutual masturbation. Eventually the listeners will abandon radio
for podcasts, MP3s, email lists to discuss the latest bands, and so
on. Radio can no longer count on its captive audience.

It never could. 45's, TV, cassettes, CDs cable, satellite TV, satellite
radio, 8-Tracks, video games, etc., etc. all compete or have tried. Radio is
pretty resilient and still reaches 93% to 94% of Americans weekly for about
the same amount of time as in 1950. There are and always have been people,
like you and Steve., who expect something else... sort of like asking for
the New Yorker to publish a Fargo edition... that is actually of interest to
nearly nobody.