[EMAIL PROTECTED] wrote:

> Gus,
>
> It would be possible to subsample any data set in order to produce any
> effect desired. Until I know how you find the subset, I can not know if that
> action is removing or obscuring the causal pattern.

Neither do I. But I need to know whether the causal pattern is still there. I
guess
you allow the possibility that the causal relation is removed. I just can't see
how.
Let's say, I take the causal relation "Smoking causes cancer". The existence of
this causal relation should not depend on how I pick my subjects, should it?
Of course, if I end up with only nonsmokers in the subsample, I won't detect
the relationship, but that does not invalidate its existence. What am I missing?

> Your requirement that I accept blanket inheritance is not intellectually
> honest.  We are dealing with numbers. We can change them by selecting them
> in different ways. I am willing to make explicit how I sample and why. You
> are not. You are acting like a child who is picking a fight in order to
> blame the victim. You want me to look unreasonable for not agreeing with
> your mysterious demands.  I ask for clarification and you say its time to
> end the conversation.  Well, you are ending it at a time that CR is winning.
> Rather than admit CR is valid, you drum up some cheap excuse to withdraw.
> That is intellectually dishonest and indicative of a poor education.
>
> Explain what you are doing and do not make highly nebulous statements
> expecting me to agree. Any pattern can be disinherited by fiddling with
> data. Define your terms, philosopher, or be dismissed with the mystics and
> sophists.
>
> And you should care, Gus.  A lot of people's lives are affected by
> statistics. It is not just a game. Its real.
>
> Bill

.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to