Simon Child wrote:
> Trefor Roscoe wrote:
>
>> Agree that its is not perfect, but it will be a darned
>> site larger
>
> OK, I'll give you that one :)
>
>> and more accurate than anything we have had before
>
> But I still wonder about accuracy.
>
> Is there any reason to believe that this new information (based on
> thousands of practices with varying degreess of accuracy and
> completeness, and in some cases with differing interpretations of
> criteria) will be more accurate than existing statistics which are
> based on probably more accurately measuring of smaller samples and
> then scaling up?
The only way to answer this would be the sort of study that Julia
Hiipersley Cox et al did in Nottingham on comparing paper and computer
data but on a much larger scale and prospectively over the next year or
two.
Unfortunately unlikely to produce any ground breaking results.
Trefor
|