> > > A trend has the same sample size as a book; it is a 12
> > week
> > > rolling average. Every third trend becomes a book.
> >
> > The one difference, of course, is weighting the sample.
> > First and second phases of a book are dependent on what
> Arb
> > gets back in terms of diaries; the third phase is subject
> to
> > manipulation in terms of balancing the sample (efforts to
> > get more return from particular demographics).
>
> A trend tends to be pretty faithful to the data when you
> compare extraps with Phase I and Phase II data. There are
> small variances, but you have the same sample size, and
> Arbitron knows by experience very closely how many diaries
> to place for each demo, so weighting is very tiny in any
> cell (look at sample report).
>
you always have to one up don't you
> > week
> > > rolling average. Every third trend becomes a book.
> >
> > The one difference, of course, is weighting the sample.
> > First and second phases of a book are dependent on what
> Arb
> > gets back in terms of diaries; the third phase is subject
> to
> > manipulation in terms of balancing the sample (efforts to
> > get more return from particular demographics).
>
> A trend tends to be pretty faithful to the data when you
> compare extraps with Phase I and Phase II data. There are
> small variances, but you have the same sample size, and
> Arbitron knows by experience very closely how many diaries
> to place for each demo, so weighting is very tiny in any
> cell (look at sample report).
>
you always have to one up don't you