HCochet said:
I never believed in the accuracy of Arbitron ratings or the new people meters or whatever it's called. I want scientific proof (which is impossible, I guess).
Actually, it is not impossible. Statistics is the only scientific discipline where "error" is not a dirty word. Based on a number of factors like sample size vs. the size of the universe and such, the margin of error can be determined.
The principle purpose of doing ratings using a statistical sample is that this is the only way to get a useful approximation of the size and characteristics of a station's audience with a cost that is bearable by the industry.
A "census" is so expensive that even our prone-to-throw-money-at-anything government can only do one ever 10 years. Even then, it is a snapshot of one moment in time and so immense that it takes nearly a year to process the data. Radio surveys are supposed to get rapid turnaround of data and be reasonably current at any given time.
Ratings don't have to be 100% accurate. The purpose of ratings is to allow advertisers to measure the effectiveness of ad campaigns with a uniform metric. And, of course, by "effectiveness" I mean reaching the right amount of people the right number of times at the desired cost per listener.
Anything less is a guess.
Ratings are clearly labeled as "audience estimates" but they are not guesses. They are a measure of a representative sample of the population projected into the universe. The same methods of research are used by the major US marketers, from P&G to Coca Cola, to make billion dollar decisions. And the data from Arbitron is used to place a big percentage of the $15 billion or so in radio billings each year.
I used to go to sleep at night to KGO and wake up to KGO. Not anymore.
Were I to conclude that you are over 55, would I be right?
I turned 50 on Dec 25 2012. So no, you would be wrong if you thought I was over 55. Also, I had been going to sleep with KGO for the previous 30 years at least. You don't have to be old to like the old KGO.