[extropy-chat] IRAQ: Weapons pipeline to Syria

[Stephen Van_Sickle]

--- Eliezer Yudkowsky <sentience at pobox.com> wrote:

> The survey methods of Roberts et al. are better than
> any previous estimates 
> provided, which is itself startling.  

Yes they are.  Better does not mean adequate, though. 
Given the conditions on the ground, I don't find it
improbable at all that it is impossible to do a good
study.

> You
> can't just throw away that 
> estimate because you don't like it, unless you come
> up with better data. 

No, it's not a matter of not liking the result, but
not liking the confidence interval. If I presented
data like that to my advisor, she would laugh and tell
me to go back and keep trying.  No, you don't throw it
out, but do you *publish* it?  And issue press
releases?

> *However*, that figure of 100,000 is *after*
> excluding the statistical 
> outlier of Falluja, which might be taken as
> representative of other 
> extremely devastated cities not visited by the
> investigators, and which 
> would have implied a figure in the range of 300,000
> civilian casualties if 
> included into the study.  I'd have included it.  

So would I.  This was another reason I don't have much
confidence in this study.  This *is* a clear example
of throwing out the data because you don't like it.  I
suspect, as you say, it was done to make the study
more believable.  Again, my advisor would throw me out
of her office if I suggested doing this.

steve