Re: Pearson-r in SQL

On Thu, 23 Dec 2004 15:00:51 -0800, Gene Wirchenko <genew_at_mail.ocis.net> wrote:

>"-CELKO-" <jcelko212_at_earthlink.net> wrote:
>
>>Add more data to the set, drop the columns with NULLs and now I can get
>>a result. For example with four more rows:
>>
>>(3kg, 6m)
>>(5kg, 10m)
>>(7kg, 14m)
>>(1kg, 2m)
>>
>>my known values are linear. But I have those other rows of partial
>>data that should reduce my confidence that I really have (r = 1). I
>
> Why? The partial data does not say anything either way.
>
>>just don't know what the SQL products like CA-Ingres are doing with
>>this, nor what statistics would do with it.
>
> I would toss it. It is not valid data.

[...]

What if the cost of getting one measurement was, say 37000 Dollars? Even if you could get only the x result and not the desired (x, y) pair, due to some experimental mishap?

Are you really ready to throw away such valuable information just because it does not fit the formula?

Greetings
Matthias Kläy

-- 
www.kcc.ch