David Eduardo wrote:

Actually, diaries are placed using a technique based on random digit
dialers, with strict geographic controls within each market's metro.
Participants are recruited based on quotas for age, sex, ethnicity, etc.
based on Claritas quantifications of each market using root Census data and
annual updates. In today's world, this is as close as you can get to a true
random probability sample where there is no recruitment bias.

There is nothing "hand picked" about the sample. Stations can not ask to
have diaries sent to anyone. It's all random.

And the diary method is going away, as the People Meter rolls out over the
next few years. It's already in Philly and Houston, and does full electronic
measurement of a perfectly balanced sample.

I participated in an Arbitron radio survey many years ago. I was first
contacted by phone and they asked if I would be interested in keeping a
diary. I agreed and they sent the diary to me. The main problem I had
was I often don't listen to a station for more than a few seconds to see
if I like the music they're playing. If not, I move on to another
station. I don't listen to most commercials either, so that's another
reason for retuning. I do this a lot with FM. It's almost impossible to
enter this kind of listening into a survey diary. There would be
hundreds of entries in a day. Since I'm also an HF listener, I had to
enter the shortwave stations in the diary. I imagine they threw mine out
after they saw that.