'Major poll shows gun ownership in decline'

[Pahoo]

I'm going to do my own survey right here

Does anyone remember when Senator Joe Manchin, from West Virginia went on the news channels and stated that he had conducted a survey amongst his gun owners constituents and most of them were in favor of extended background checks, that he was supporting. I'm sure he didn't ask any Hatfields. .....

Be Safe !!!

[Snyper]

They only surveyed 2000 people, so it's all meaningless hype for the gullible

[Derbel McDillet]

"Guns? My wife and I both hate guns!"

[Ike666]

If you look at the GSS data over time, the response to the gun ownership question is interesting. Over the years the answer to "Do you have a gun in the home?" question has showed a steady decline. Even in recent years. What media reports like the Yahoo article don't tell you is that the proportion of respondents refusing to answer the question has steadily grown. In fact, it has grown proportionally to the decline in reported "ownership."

In essence, the percent acknowledging having a gun in the home has probably remained fairly constant since the late 80s. That is if we assume that those who refuse do so for the reasons so strongly stated by the forum membership.

In survey research a sample size of 2000 is actually pretty substantial. NORC, who conducts the GSS, uses a national probability sampling methodology. This means, statistically, it is matched on demographics to the U.S. population. The sample size gives it more than adequate statistical power.

[Tom Servo]

Except do those statistics tell us anything about how many of those sales were to people who already had a gun, and how many to people who were acquiring their first gun?

They don't, but the burden of proof isn't on us. It's on the people making the allegation. The data is enough to cast serious doubt on their claims.

I wish I could snark half as well as she does.

Nobody can. She's the master, and I can only bow to her superior snark.

[Frank Ettin]

Originally Posted by Tom Servo

They don't, but the burden of proof isn't on us. It's on the people making the allegation....

True enough. But I think we have an interest in knowing for our purposes. First, we don't want to become complacent. Second, if we had solid data to rebut the anti-gunners, it would be nice to be able for us to demonstrate that their claims illustrate their propensity to make purposefully misrepresentations.

[SaxonPig]

I'm sure many people didn't reveal gun ownership on the phone to "someone from the government." The soaring sales on guns and ammo the past 6 years plus the burgeoning interest in firearms by hoards of new women shooters puts this survey in the trash can. They are lying or they were lied to by reluctant respondents.

[oldcabin]

If no one actually owns or buys any guns, then why can't I find any 22 on the shelf anymore

[egor20]

My dog ate my homework and my guns.

All they would get from me is a dialtone.

[Tom Servo]

But I think we have an interest in knowing for our purposes.

Oh, absolutely. I don't know how we'd gather it any more effectively than the antis would, though.

[Webleymkv]

I would be very interested to see the exact methodology of the survey because, unless there's a lot more to it than the article implies, it sounds seriously flawed. From the article:

The General Social Survey is administered by NORC at the University of Chicago, primarily using in-person interviewing. The GSS started in 1972 and completed its 30th round in 2014. The typical sample size was 1,500 prior to 1994, but increased to 2,700-3,000 until 2008, and decreased to 2,000 for the most recent surveys. Resulting margins of error are between plus or minus 3.1 percentage points for the smaller sample sizes and plus or minus 2.2 percentage points for the larger sample sizes at the 95 percent confidence level. The 2014 survey was conducted March 31-Oct. 11, 2014, among 2,538 American adults. The GSS 1972-2014 Cumulative File was used to produce the statistics presented.

So the survey was of roughly 2500 people and done at the University of Chicago. This implies that the interviewing was done in Chicago and that the majority of people interviewed were likely from Chicago or the surrounding area. I don't think that I need to tell anyone here that Chicago is not gun friendly in the least and that, in all likelihood, attitudes about guns probably won't be a positive there as they would be in another city like Dallas, Phoenix, or Indianapolis.

Also, because of the draconian gun laws of Chicago, there is a pretty good likelihood that a good number of interviewees lied about owning a gun for fear that they may have run afoul of the law either knowingly or unintentionally. It is pretty widely accepted that gun ownership in places with very stringent gun laws like Chicago, New York, or Washington D.C. is often grossly under-reported because a large percentage of the guns are owned illegally. As a good friend of mine who grew up in Brooklyn once explained, there's a lot of guns in NYC, people just tuck them away and keep quiet about it.

Now, maybe the methodology of the poll is different than what I'm seeing in the linked article, but I went to the NORC website and still cannot find the original data.

[44 AMP]

I spent decades working in a chemical industry, and I understand very well the concept of a representative sample.

I won't claim to understand the arcana of statistical analysis, but I have always had my doubts about how any poll (sample) of a thousand or two thousand people can actually be accurately representative of the over 300 MILLION people in this country.

To get a good ONE MILLILITER sample of a 5,000 gallon tank, it takes at least a calculated time of agitation (mixing/blending tank contents to ensure a homogenous mixture) which can be hours. A 100,000 tank can take days.

I'd be willing to bet that if I conducted "personal interviews" with 2500 people entering the SHOT show, I'd get a MUCH different result than the people in Chicago.

I got a phone call the other day from some one with a heavy accent, claiming they were the IRS, and I was going to be sued for not paying my taxes. When I asked them what my name was, they said "Your name is (obscenity) and hung up.

NO WAY I am going to tell anyone on the phone ANYTHING about my firearms ownership, or sad lack thereof...not because I worry, because I consider it a matte of principle. I do talk about what I have, here, and don't care who reads it. Perhaps I'm being inconsistent, but it won't be the only thing in my life that fits that description.

[Webleymkv]

Originally posted by 44 AMP
I won't claim to understand the arcana of statistical analysis, but I have always had my doubts about how any poll (sample) of a thousand or two thousand people can actually be accurately representative of the over 300 MILLION people in this country.

That's an excellent point, 2,500 people represents approximately 0.0008% of the total population of the country . For that matter, 2,500 people only represents about 0.05% of the population of Cook County, IL alone (the population of Cook County was slightly over 5,000,000 per the 2010 Census).

To my mind, in order to get a representative sample of the country at large, you'd have to survey, at the very least, several million people. Not only that, but we're lacking demographic information about the interviewees such as age, gender, race, ancestry, marital status, religious affiliation, sexual orientation, etc. all of which could play a part in people's answers.

[ballardw]

The organization conducting the poll is so proud of this work you cannot find it on their website: http://www.apnorc.org.

I have worked with survey data for the past 16 years. If a project doesn't provide some good description of how the sample was designed then that is one big flag about the results. The second is if you cannot find the actual questions asked and response categories then be very cautious with the results.

With a supposed "in person" interview I would be very interested in knowing 1) where the offices of the interviewers were and 2) what is the furthest distance from any of those offices that surveys were conducted. If the surveys are all within 30 miles of New York, Chicago, Seattle, Portland, San Francisco, Los Angeles and such then it really shouldn't be claiming much about nationwide.

[spacecoast]

The survey taker would have to be holding me at gunpoint to get me to talk. Of course, that could be risky.

[Ike666]

It really isn't that hard to find.

publicdata.norc.org:41000/gss/Documents/Codebook/A.pdf

It's not the messenger. The sample is a national probability sample, spread around the country. Statistically it will do a better job of predicting the national population than any other survey design. It is sure a lot better than anecdote and gut feel.

That said, I believe it is inaccurate in this case. Not because of flawed sampling, but for the reason so many here on TFL cite. Since 1972 fewer and fewer people have been willing to provide any answer to that question (and it is just one question about gun ownership).

I think most of the refusals are probably owners but I have no data to support that. In point of fact the closest thing we have is the NSSF NICS data. And it is an analog at best. Their formula attempts to extract the permit checks, transfers, and pawn remits. So the NSSF data are an approximation of new sales absent other administrative checks. But it assumes purchases/transfers are equivalent to ownership.

The only truly knowledgeable body is the industry itself. They know their own sales. They know how many are exported and how many are government sales. But their data are proprietary and they are disinclined to share.

So, if we actually want to know the answer we have to be willing to cooperate. Absent that it is all uniformed speculation.

[Pawatch]

In California no less

EUREKA, Calif. -
Handgun sales in the state of California reached a record high in 2014, with more than 500,000 sold, doubling the amount of handguns sold in 2010, according to the State of California Department of Justice.

http://m.krcrtv.com/handgun-sales-in...-2014/31662816

[dcludwig]

I once owned some guns, then the fire of '98 destroyed them all. *sniff*

[Skans]

Gauging gun ownership based on gun sales is also flawed. We all know that a gun owner is more likely to purchase another gun than a non-gun owner is to purchase his/her first gun. Still, I believe the NRA tracks gun ownership to the extent that it can, and I'd trust its estimates over any survey.

[Son of Liberty]

Guns? No habla, gringo.

[Tom Servo]

We all know that a gun owner is more likely to purchase another gun than a non-gun owner is to purchase his/her first gun.

Do we? 80% of my rifle sales from 12/2012-04/2013 were to first-time buyers.

[DaleA]

80% of my rifle sales from 12/2012-04/2013 were to first-time buyers.

I don't dispute your numbers but how do you know they were first-time buyers?

FYI: Waaaaaaaay back when a journalism course I took run by a professor that had a 'thing' for statistics showed us that if the sample were truly random then about 1600 was all you really had to 'poll' or 'sample' to be pretty sure of the results. That is, a company making light bulbs (all made exactly the same of course) could test a random sampling of 1600 bulbs and then tell you pretty closely how long they would last and what chance you had of getting a bad bulb would be.

Obviously you cannot 'survey' folk at the SHOT show for firearms stuff any more than you could survey NPR members for gun stuff. You have to have a representative sample. Getting a 'representative' sample is where the expertise comes in and where the polling companies show whose the top dog and who is the also ran.

See Jimmy Stuart in the movie 'Magic Town' for more polling expertise.

[Ozzieman]

Joseph Goebbels Quote

“If you tell a lie big enough and keep repeating it, people will eventually come to believe it. The lie can be maintained only for such time as the State can shield the people from the political, economic and/or military consequences of the lie. It thus becomes vitally important for the State to use all of its powers to repress dissent, for the truth is the mortal enemy of the lie, and thus by extension, the truth is the greatest enemy of the State.”

[rickyrick]

Well since records aren't supposed to be kept... All buyers are first time buyers lol...

I shot my AR today... I'm so ashamed and I went home feeling dirty

[Tom Servo]

I don't dispute your numbers but how do you know they were first-time buyers?

If you maintain your records in digital form, it's an easy thing to check. The majority of rifle buyers during that period had no prior history with us.

Does that constitute a representative sample? No, but it does cast some doubt on the idea that the panic buying was largely existing gun owners who decided to augment their collections.

(The folks paying 300-400% premiums on the secondary market certainly weren't informed members of the gun culture.)