|
The problem with creating a statistically significant sample is that the show can't know you're gaming them, or it skews everything. I think the only approach that would make sense would be to listen to the show a lot and figure the percentage of callers who disagree with the host, although then you can't know whether you're measuring host authoritarianism or audience demographic.
It's interesting though, even though it's far from scientific. It's been demonstrated before that liberals and conservatives consume media in very different ways. Liberals tend to want all the information, while conservatives tend to only watch or listen to things they already agree with. I can't remember where I read that, but there was a big study done that showed that. That's why Fox News is so popular with conservatives and everything that's not Fox News is accused of being the "liberal media".
|