>I tend to be very interested in results that are "discarded" or "excluded",

> as they would tend to be data points that put the work in an entirely
> different light if they are addressed as valid results.  One does not
> discard data points without very clear criteria, in my view.
>

I'm in complete agreement with you on this point Jim.  I often search out
the original data if possible to see just what was discarded, and why.

One thing that I notice in research is that there is often high initial
mortality in bees that have been marked or treated.  I get the feeling that
some researchers (or their students who actually do the work) are simply
not comfortable with handling bees, and inadvertently stress or injure them.

For example, at least two studies found a very high mortality rate in
paint-marked newly emerged bees when returned to the colony.  Contrarily, I
paint marked 600+ newly-emerged bees a couple of weeks ago, but did not
find a single dead one in front of the hive the next morning.  But I'm
quite used to handling and marking individual bees.

Obviously, if the test and control group each receive the same stressful
handling, then that identical treatment would theoretically not skew the
results.  But then the paper should read "The Effects of xxx Upon Bees
Stressed by Handling."  My point being that in order to get real world
results, we should be working with real world bees--not those traumatized
by rough treatment.

-- 
Randy Oliver
Grass Valley, CA
www.ScientificBeekeeping.com

             ***********************************************
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