Hi all

As I said in a previous post, surveys are not the same as controlled trials. In fact, here is some more explanation from Katie Lee:

> Survey data can be a big, dirty mess, meaning that there are all sorts of biases that can be introduced into the data, like relying on beekeepers to correctly recall practices or a bias in the demographics of the respondents. Because of this, surveys are one of the worst ways to get at causation. We ask so many questions that there could be an effect that look significant, but could be really just do to chance. 

> For example, we found that beekeepers that replace 50% of their comb every year tend to have higher losses. Now, it could be that old comb confers some type of microbial advantage to the bees, it could be that the bees spend too much energy in building comb, it could be due to something linked to the beekeepers like they tend to be mostly new hobbyists, or it could just be due to chance.

* In other words, the whole purpose of a controlled trial is to eliminate all these other things so you are only looking at the things you want to be looking at. But alas, the more controlled a trial is, the less it's likely to reflect the real world. Somehow, we have to tease these things out by comparing different reports from different areas. That is one of the many benefits of an international discussion such as this.

PLB

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