United States & Gun Control discussion.

Deathy McDeath said:
Not necessarily. You can design a nationally-representative survey with few respondents as long as your survey is designed well, and you achieve a strong response rate. You can actually check the math here: Sample size calculator - CheckMarket
For example, I entered 300,000,000 for the population size (just to get a nice round number. past a certain point, population size doesn't really mean much), a response rate of 35% (the only thing that a higher response rate achieves is increasing the number of people invited to the survey), with an MoE of 2% and 99% confidence interval. That resulted in about 12,000 people people being surveyed, and a sample size of 4147 (again, response rate 35%).

It is, however, a little difficult to tell exactly how they structured the survey and what the response rate was because it looks like they haven't published the study yet. I'd be interested to see what their methodology was. I'm also curious as to some of the claims in the article. For instance, it said that the survey sampled approximately 4,000 gun owners, but it also stated that, "The study found that 22 percent of American adults say they personally own a firearm. " I'm not sure whether there was a distinction made between having a firearm and OWNING a firearm (e.g. having a gun in the home vs. personally owning one), or if that's an oversight by the writer. Regardless, if 4,000 gun owners were included in the sample, I don't see how you could draw out the 22% ownership rate from that. I can definitely understand the line about 3% of gun owners having half of the guns, but I don't get the 22% figure.

Again, I'd like to see the survey itself before drawing conclusions about the sampling method, but the statistical reasoning has the potential to be sound.
Click to expand...


Can you repeat that for those of us on crack? :thumbsup:
 
Ocoka One said:
Can you repeat that for those of us on crack? :thumbsup:
Click to expand...

You don't have to have a colossal sample size for an accurate sample, if you set it up correctly (to accurately represent the population).

I have some strong doubts that they did so in any meaningful way.
 
Ooh-Rah said:
According to the results of a Harvard-Northeastern survey of 4,000 gun owners, just 3 percent of American adults own half of the nation’s firearms, according to t.

I say horse shit, this is nothing more than the continued media spin to show that gun owners are a minority segment of America and paint them as paranoid.

Just 3 percent of American adults own half of the nation’s firearms, according to the results of a Harvard-Northeastern survey of 4,000 gun owners.

The survey’s findings support other research showing that as the overall rate of gun ownership has declined, the number of firearms in circulation has skyrocketed. The implication is that there are more guns in fewer hands than ever before. The top 3 percent of American adults own, on average, 17 guns apiece, according to the survey’s estimates.

The survey is particularly useful to researchers because it asked respondents not just whether they own guns, but how many and what types of guns they own. This makes for one of the clearest pictures yet of American gun ownership.

The bolded immediately shows the survey flaws.
1) I'm not likely to tell anyone doing a phone survey if I own guns
2) There is ZERO chance I am going to tell them how many and what types I own. Did they as for the location and combo to the safe next?

Three percent of American adults own half the nation's guns
Click to expand...

I believe it. Think of how many people own a dhotload of guns, and how many own zero. It is still a shitload of people with 1-2 guns.


ThunderHorse said:
10,000 respondents is the accepted number for scholarly studies.
Click to expand...

Ugh what? Did you literally just pull that out of your ass?
 
A 10000 person sample would need something like 30-50k people surveyed. That would be soooooooo expensive.
 
I do love when a study comes out how everyone including me has a fucking Ph.D in stats.... we are all like "the p-value...."
 
TLDR20 said:
A 10000 person sample would need something like 30-50k people surveyed. That would be soooooooo expensive.
Click to expand...

Not to mention the colossal waste of time and effort involved in gathering data after you have all the data you need and your study has sufficient statistical power.
 
Marauder06 said:
No matter what the sample size is, I just skew the results to reflect whatever pre-determined political view I had that drove me to undertake the study in the first place. Sample size of 10, or 10,000, or 10 million. No drama.
Click to expand...

Entering a study and results analysis with a distinct personal bias.

The Trump Science Methodology has officially begun it's spread. :D;-)
 
DocIllinois said:
Entering a study and results analysis with a distinct personal bias.
Click to expand...

Every time I get a call or someone at my door with a "survey" or some such, I always give answers that don't reflect my "truth". Certainly I am not the only one who does this.

Last night we had someone at the door from the League of Conservation Voters. They asked who I was voting for. I said "not Clinton." She said, "So Trump." I said, "I didn't say that." She said, "well, who else is there?" So I said, "good day, madam."

They have their game, I have mine.
 
I think the survey is crap.

Florida alone has issued--as of March 2015--1.4-million Concealed Weapons Licenses. The GAO estimates there are at least 8-million active concealed firearm carry permits in the US. Pennsylvania has issued 1.1-million since 2008. And twelve states have fully unrestricted carry laws.

Obviously, any toll of CCW permits doesn't account for the millions of legally owned firearms whose owners don't possess a CCW. I know people who have only a shotgun or a 22. What about people who got a CCW and only have one handgun, the one they carry?

This bunch of shithot rocket aces up there at Harvard could've started with published data before hitting the phones.
 
Ocoka One said:
I think the survey is crap.

Florida alone has issued--as of March 2015--1.4-million Concealed Weapons Licenses. The GAO estimates there are at least 8-million active concealed firearm carry permits in the US. Pennsylvania has issued 1.1-million since 2008. And twelve states have fully unrestricted carry laws.

Obviously, any toll of CCW permits doesn't account for the millions of legally owned firearms whose owners don't possess a CCW. I know people who have only a shotgun or a 22. What about people who got a CCW and only have one handgun, the one they carry?

This bunch of shithot rocket aces up there at Harvard could've started with published data before hitting the phones.
Click to expand...

I like to think of this 'study' or poll with this analogy:

Conduct a similar poll, but instead of guns, use "silverware." Now you tell me how can you conclude the real ratio of owners to silverware. I mean, sure, everyone has at least one of each: one fork, one spoon, one knife. How many have more than one? Did they purchase them from Bed, Bath, & Beyond, or at a yard sale or an estate sale? Is it fine silver that is boxed up, or your average Wal-Mart cutlery that you would use every day?

Sure, you don't need a permit to purchase; but nor do you need a permit to purchase a rifle, and if you get the rifle in a private face-to-face transaction, you don't need much at all.

There are far too many holes in this to be respectable, and far too hard to generalize.
 
And where do they get the "rate of gun ownership has declined"? How do they figure that?

I could understand a finding that says gun sales have declined. After the hysterical gun & ammo buying frenzy in the first few years of the Obama administration, I wouldn't be surprised to see a fall off.
 
The funny thing about surveys and the like, is how do you trust them? I've quoted the FBI's data on crime and we all know it to be wrong, yet a study will come along and say the exact opposite of the FBI's crime data where guns are concerned. What if they are both wrong, then what?

That's right, we don't freaking know because we don't have the God's Eye view into the home of every American.

Besides, we're running studies against a Constitutional right in the hopes we can keep/ rewrite said right? (Sorry for the rhyme or whatever) Unbelievable.
 
Devildoc said:
Every time I get a call or someone at my door with a "survey" or some such, I always give answers that don't reflect my "truth". Certainly I am not the only one who does this.

Last night we had someone at the door from the League of Conservation Voters. They asked who I was voting for. I said "not Clinton." She said, "So Trump." I said, "I didn't say that." She said, "well, who else is there?" So I said, "good day, madam."

They have their game, I have mine.
Click to expand...

That is a pretty obstinate thing to do. Why not just be honest...?
 

Similar threads

Devildoc
House discussing citizen-only voting bill...
Replies
19
Views
381
The Quartermaster
The Quartermaster
Marauder06
1911 Questions--Barrel and Slide Swap
Replies
14
Views
779
Diamondback 2/2
Diamondback 2/2
R.Caerbannog
Distortion of History
2
Replies
32
Views
569
R.Caerbannog
R.Caerbannog
Dame
F Around and Find Out
Replies
2
Views
477
Chopstick
Chopstick
AWP
Computer Help
Replies
4
Views
90
RackMaster
RackMaster
Back
Top