Join Bridge Winners
The problem with polling

I've been following a long thread (not bothering to link) regarding the resolution of a misinformation case.  Not to put words in anybody's mouths, but roughly, the two sides come down to:

A.  The directors polled a group of peers of the players affected and found no damage, and

B.  There's obviously damage, given that the NOS was adversely affected.


And there's lots of discussion about whether the NOS's failure to bid game was a result of the MI or simply subsequent to it.  

And the conversation goes around in circles.


Here's the problem, as I see it.

The fact that a question is being asked nearly always skews the answer.

It's often unclear which way the answer will be skewed, but it happens nonetheless.  The people being polled do not naturally think exactly as they would at the table, essentially because they are "warned" by the fact that they're being polled that something strange is going on.

In some cases this can be adjusted for in the design of the poll.  For example, in a hesitation case, you don't tell the pollees about the hesitation.  But not every case can be managed that way, and there's still the basic problem--if you bring me a special question, I'm going to treat it as if it's special.


I don't know how to solve this, but then, there are plenty of people here who might reasonably have better ideas than me...


So, what do you all think?

Getting Comments... loading...

Bottom Home Top