"Bart Bailey" wrote in message
...
In posted on Thu,
15 Nov 2007 21:42:05 -0800, David Eduardo wrote: Begin
"Bart Bailey" wrote in message
...
In
posted on Thu, 15 Nov 2007 04:49:19 -0800 (PST),
wrote: Begin
ALL the people in Lake Woebegone listen to The Radio. . .
But how many of them bother to participate in
Arbitron, Scarborough, Nielson or whomever's surveys?
Nielsen is TV. Scarborough is an extension of Arbitron, who does radio
research.
Just a load or nitpicking to avoid my point that the number of
respondents to any survey is questionable, especially in the very busy
and preoccupied target 18-34 demographic, which tends to nullify any
significance paid to those numbers.
The only place this is an issue is the New York PPM "trial data" phase,
which is not currency anyway. In the other 280 markets, there is no issue.
The number of respondents "to any survey" is hardly questionable. One
designs a poll based on the economic inability to sample everyone in a
census, and picks a sample size where the margin of error is acceptable in
relation to the cost. The samples in Arbitron and Nielsen are
error-acceptable and cost-justifiable for the stated purpose, which is
placing a value on broadcast time for advertisers. A few percent error does
not affect usability... this is not a political race where one person wins
and everyone else loses. In large markets, dozens of stations "win" by
having a metric to establish pricing.