> > Good points.
> >
> >
> > >
> > > I'm curious, how does going up 1/10th of a point
> represent
> > a
> > > big win? Also, how could the ratings reflect any
> > > "migration" of former KIZS listeners when this trend
> > > represented listening in August, September and October?
> > >
> >
> Was looking at the individual month of Oct. not the rolling.
> Afterall, it will be in the Fall book and last time I
> checked August and September won't be.
>
You can't read anything into individual months. If they're extrapolations, they are wildly innacurate. If they're the monthlies from Arbitron, they are unweighted and more importantly, aren't representative of the sample. For example, say Arbitron must get 200 diaries from 18-34 males to be within their projected sample for the fall book. They won't get 67 in October, 67 in November and 66 in December. They take what they can get. If they get all 200 in October and November, they stop recruiting 18-34 males, scheduling ones they come across in their recruiting of other demos to participate in the next survey. If they're short on 18-34 males in October and November, they will pull out all the stops (shudder) to make their number in the month of December. It's not till you get the full survey that you approach any sort of statistical reliability.
Besides, it takes months for a competitor's going away to reflect in a station's ratings. I guarantee you if you check the diary comments a year from now, you'll still see people writing down "92.1 Kiss FM". It sucks, but perception is reality; that's what you live and die by in radio.