On Tuesday, June 20, 2017 at 1:35:29 PM UTC-4, Rudy Canoza wrote:
On 6/20/2017 4:23 AM, Jim Wilkins wrote:
"Just Wondering" wrote in message
...
On 6/19/2017 11:42 PM, tRudy Crayola wrote:
On 6/19/2017 2:57 AM, Just Wondering wrote:
On 6/18/2017 2:39 PM, Preston Hamblin wrote:

The percentage of households that own guns is under 30%.

Prove it - or admit you're posting bull****.

The percentage of Households that will admit to owning guns is 30%.
The rest will merely blow your ass away if you arrive uninvited
through a window.
The only data I have ever seen, which is bull**** anyway for the
reason
you implied (admitting to owning a gun is not the same as owning a
gun), is a 2014 survey, and the percentage then admitted to was 32%.
Given the record gun sales in the three years since that survey was
taken, the percentage has surely risen since then.

I'm still waiting for Hamblin to prove his assertion.
. . .
. . .
OK, I'm done waiting. Hamblin's silence is an admission that his
assertion is bull****.

It's no secret that the people who go to the trouble of surveying if
households have firearms are marking them for search and confiscation,
since their other attempts at registration have been frustrated. If
they really cared "for the children" they'd be fighting drugs instead
of trying to legalize them.

Who among you would be dumb enough to tell a stranger if you keep gold
in the house and how it's secured?

This is the wrong question. The correct question - actually, questions
- a

1. Do gun owners respond to the survey at different rates today than
they did in the past?
2. If they do respond, have they begun responding untruthfully at
higher rates now than in the past?

There is no sound reason to believe the answer to either question is
"yes". Gun owners probably respond at the same rate as in the past, and
those who respond probably respond truthfully at the same rate.

What's measurable and measured, without the benefit of experimental testing, is that ALL types of respondents respond at lower rates than in the past. This may bias responses of gun owners because it's clear that conservatives are more likely NOT to respond than they were in the past than liberals are.

But, as I said, some experimental testing has shown that people who would not tell the truth on a gun-ownership question are the same people who don't respond to the survey at all, and thus are not counted either way. If the survey is a good one, it will include a couple of political questions and then weight results to compensate for non-respondent bias based on political leanings.

What may not be understood here is that surveys run by the big organizations -- Gallup, Pew, or the GSS. for example -- take account of these kinds of problems and make adjustments to get valid statistics. Lying on a question about gun ownership is part of an entire class of survey problems called "Demand Characteristics." You may remember one conducted a few years ago in which the question was asked of known CCW holders. The percentage of false responses was extremely low.

The corrections are good, for the most past. Look at the result of the last election, which was LOADED with demand characteristics. The major pollsters' average was within something like 1.2% of the actual results: spectacularly close for any survey with an overall result close to 50%. That's the most difficult scenario and the polling results don't get any better.

State polls, which were performed mostly by smaller pollsters and with much less professionalism, were much worse. That's to be expected.

--
Ed Huntress