On Thu, 23 Nov 2017 01:07:34 -0000 (UTC), Przemek Klosowski
wrote:

On Sun, 19 Nov 2017 10:35:21 -0500, Ed Huntress wrote:


That only proves CCW holders don't mind revealing information the
government already possesses, and says -nothing- about other people.
Your logic is unusually sloppy here.

"My" logic? By what logic do you conclude that CCW holders tell the
truth,
while other gun owners are liars?

There isn't any. This was a double-blind test -- neither the pollster
nor the subject knew that they were chosen from a CCW database

Jim has a point that CCWs are a biased sample.

There is a potential for bias. Further comments in this thread
indicate that there probably wasn't any actual bias.

Try this
Gedankenexperimant: what if they asked convicted felons instead of CCWs?
They would have to answer NO, regardless of whether they have the gun at
home or not.

Convicted felons have nothing to do with the research results we were
talking about. The bias there is based on a risk of being jailed. The
incentive there is 'way over the top.


Having said that, I suspect that CCWs are much more consistent on this
point with the general population than CFs

No doubt. g

When social-science researchers look for potential bias, and try to
design tests to measure it, they separate cases where there is an
obvious or likely incentive for bias, from those that have no such
known motive or incentive. Very often, the "obvious" bias turns out to
be nonexistent, in terms of results. The problems tend to fall on
those that indicate a bias, and a correlation that goes with the bias,
but no obvious reason for it.

That can make survey designers tear their hair out. But if they can
measure and predict it, it doesn't affect the results.

It's a neat and satisfying field for very smart people who have good
insights. I used to live a few blocks from the Gallup and Public
Opinion Research offices in Princeton, NJ, and the sons and daughters
of workers there, who were my high school classmates, were some of the
brainiest people I've ever known.

--
Ed Huntress