> No.  It won't establish "credibility", rather, 
> it will establish a database of dosages at 
> which either harm, or no harm, was shown.

It will establish credibility by painting a consistent picture, or undercut
it by painting a random and inconsistent one.

And the dosages would be more accurately labeled as those at which harm or
no harm was CLAIMED, not "shown", as we have seen many claims of harm that
could not be replicated, or were easily, and thus were more likely due to
"methods" than "toxins".  On the other side of the coin, it is also very
possible that harm was done to bees but not noticed in the context of the
tests done.

> Poorly done studies are more the exception than the rule

In general, this is very true, but in the specific case of pesticide bee
toxicity studies coming out of Europe, we have time and again seen methods
used that have evinced massive ignorance of bees, for example the classis
error of capturing bees as they emerge from the hive to use in "homing"
studies rather than returning foragers with pollen on their legs.  The
difference is that many emerging bees may be young bees, those that have not
yet completed their orientation flights, and thereby unable to return
promptly (or at all) from the release point to the hive.  We've also seen
blatant advocacy and preconceived notions in many papers.   

> ...and we should ensure that multiple points of view are taken seriously.

I would rather let the data do the talking.  "Points of view" not backed by
hard data are "speculation" to me.

> Will we see a scattergram when we plot dose versus % survival from
multiple studies?  
> We expect so, and in that case, what is the next correlate we would impose
on the data?

Well, we had better agree on rules of engagement before we start, as making
the rules after we see the plots would be putting our thumb on the scale.  I
suggested that data points near each other in the scattergram should be
looked at to see if that general area of dose and exposure time resulted in
"similar tangible results", as similar observations would tend to enhance
the credibility of that subset of the data points. 

Very different results from nearby points would tend to undercut the
credibility of the cluster of points, as they did not support each other
with similar results.   Regardless of the "tightness of the pattern",
similar effects from similar doses is where the rubber meets the road.


And now for a totally dispassionate and impartial observation before I go to
dinner:
J!... E!... T!... S!...  JETS! JETS! JETS!

             ***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software.  For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html