[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: A possible GFDL compromise: a proposal

On 2003-09-27 12:37:52 +0100 Mathieu Roy <yeupou@gnu.org> wrote:

You must try to avoid bias when designing the data collection

This disagrees with your earlier comment.

What is called here "controlling for bias" is indeed introducing
bias -- a big one.

I did not defend it. Please try not to continue arguing points already made. It makes for long and tedious exchanges.

Avoiding bias means trying to collect _raw_ data.

Cobblers does it.

When you question
someone, he should not be able to clearly know what answer you expect
from him.


, else you may alias different factors and
The biggest factor of bias here is the author (of the question) point
of view.

I disagree and say that it's impossible to know, but that is moot.

waste a lot of money.
It is about money here? Why talking about money here?

It is usual to talk about data collection being "cheap" or "expensive" and use an equation of money for effort even when it is not paid. I have even seen that in French-language statistics texts.

MJR/slef     My Opinion Only and possibly not of any group I know.
http://mjr.towers.org.uk/ gopher://g.towers.org.uk/ slef@jabber.at
 Creative copyleft computing services via http://www.ttllp.co.uk/

Reply to: