Marylanders strongly support gun control proposals, poll finds

[erwos]

How do you know 800 is enough to form a statistically relevant percent? I have no idea, but after reading this, I suspect that it may not be enough.

Nope, it's enough. 800 random people out of the population is a pretty reasonable sample size. I am going to guess you don't understand the link you posted, which pretty much shows the opposite of what you're saying. SAMPLE SIZE DOES NOT NEED TO BE PROPORTIONAL TO POPULATION SIZE.

These guys are a bit liberal, but it's a good read on why huge sample sizes are not really as helpful as some people think:
http://www.publicagenda.org/pages/best-estimates-guide-sample-size-and-margin-error

Loaded questions, skewed samples, etc. are different problems. The number of people polled is not. We ask gun controllers to not be ignorant, let's not show that same ignorance of basic statistics on our side, please.

 

[Tyeraxus]

How do you know 800 is enough to form a statistically relevant percent? I have no idea, but after reading this, I suspect that it may not be enough.

Simple statistics. 800 people works out to be right at +/-3.5% for a population of 5.8million and a confidence interval of 95%. If the sample isn't truly representative that would be a sampling error, but the size itself the sample is pretty spot on. I'd be more interested in seeing the question asked and the response choices available. Waaaaay easier to lead with the questions than to play games with the numbers afterwards.

 

[EL1227]

Right on, Right on, Right on ...

Smokey hit that like the fist of god in the comments section

http://www.mdshooters.com/showthread.php?t=97698

View attachment 73674

I added a left hook to your (right) fist of god ...

 

[smokey]

I added a left hook to your (right) fist of god ...

I see that. And here comes an uppercut...

 

[EL1227]

And the progressive liberal cries foul

I see that. And here comes an uppercut...

No fair, we're double-teaming him. We're a couple of schoolyard bullies. Waaaaaaa ....

 

[smokey]

It's okay to pass any law you want that violates the constitution, since it can always possibly be found unconstitutional a few years later

 

[Haides]

It's okay to pass any law you want that violates the constitution, since it can always possibly be found unconstitutional a few years later

Don't tell me that fool just said that...

 

[R357]

Simple statistics. 800 people works out to be right at +/-3.5% for a population of 5.8million and a confidence interval of 95%. If the sample isn't truly representative that would be a sampling error, but the size itself the sample is pretty spot on. I'd be more interested in seeing the question asked and the response choices available. Waaaaay easier to lead with the questions than to play games with the numbers afterwards.

Wait a second there sparky. 800 out of a 5.8 million population represents 0.00013793% of the population.
Not reflective of a 3.5+/-% of the population nor does it speak to the assumptions you are relying on for the presented confidence interval especially in view of previous data collected regarding magazine capacity.

 

[smokey]

Don't tell me that fool just said that...

yup silly people abound in this world.

 

[alucard0822]

Wait a second there sparky. 800 out of a 5.8 million population represents 0.00013793% of the population.
Not reflective of a 3.5+/-% of the population nor does it speak to the assumptions you are relying on for the presented confidence interval especially in view of previous data collected regarding magazine capacity.

The number sampled isn't the problem, you could get reasonably accurate results with a quarter of the number they sampled. The problem is that the group that did the polling is owned by Michael Bloomberg, and his extreme bias is well known. The demographics are also important, polling in a disproportionatley urban area, age group, or any other numbers of areas can skew the results, then there are the questions themselves and their wording, even then sometimes the masses are asses and the results are true but crappy. Their source of the numbers to call, even the time they called make a difference, polls are easy to manipulate, that is why they are a favorite of anyone with an agenda, and why our founders established a constitutional republic to protect rights however unpopular, and prevent mob rule by the majority.

 

[Mark75H]

Simple statistics. 800 people works out to be right at +/-3.5% for a population of 5.8million and a confidence interval of 95%.

Wait a second there sparky. 800 out of a 5.8 million population represents 0.00013793% of the population.

203,000 is 3.5% of 5.8 million, but a (truly random) sampling of as few as 800 will give the correct result 19 out of 20 times ... That's what Ty is saying

Knowing the exact question and allowed answers is the key ... also knowing the given "facts" stated before the question is as important.

We don't know what the people were told or asked, but it prompted them to give what we truly suspect is a erroneous answer compared to giving them real knowledge and real choice in the answer ... or the sample was not truly "random"

 

[Bolts Rock]

Enthusiast jump took a dump on their claims.

 

[AliasNeo07]

In a proper random sample, 800 people is MORE than enough to have a statistically valid conclusion. So, let's lay off the "OMG not enough people polled" whining, because it's flat-out wrong as a criticism of the poll. If it wasn't random, that's a much different issue... but I suspect it was.

Citing an internet poll as "contradictory evidence" is laughable. They're not scientific in any way, shape, or form. Hell, we've all seen how we jump on those... the people answering them are NOT representative of the population as a whole.

I was honestly surprised it was only 60% in favor of an AWB, truth be told. That's close enough that a motivated minority actually could derail it. The mag limit, ugh, that's not looking good.

I'm going with erwos on this. Why does it shock you guys to think a majority of Marylanders would support an AWB? The poll may very well have been accurate. This isnt TN...it's MD.

It wouldnt be the first time a majority of people are on the wrong side of an issue and won't be the last.

 

[Mobile]

The only functional difference between these guns and the non-evil gun is the magazine capacity, and that's already capped at 20 in MD.

But that won't make a difference to them I suppose

 

[smokey]

looks like the anti-trolls are taking turns in the comments section. It's was elcid's turn and now I think mojosfun is going for the "king dumbass" crown

 

[R357]

203,000 is 3.5% of 5.8 million, but a (truly random) sampling of as few as 800 will give the correct result 19 out of 20 times ... That's what Ty is saying

Knowing the exact question and allowed answers is the key ... also knowing the given "facts" stated before the question is as important.

We don't know what the people were told or asked, but it prompted them to give what we truly suspect is a erroneous answer compared to giving them real knowledge and real choice in the answer ... or the sample was not truly "random"

First I have little doubt that polls can be skewed by the questions, language, context and even tone they are presented in. Further steps can be easily taken to knowing skew these polls through targeting respondants (geographically, religious associations, or even time of day if handled via telephone implemented polls).. So lets just say there are literally tons of ways to skew the data and agree on that point.

My issue here is that the sample size is no where near the 3.5% percentage indicated by Tyeraxus post nor does 0.00013793% of a population (800 of 5.8mil) guarantee of confidence interval of 95%as implied.

Confidence intervals are not simply proportional to the number respondants or even percentage at this low of a level. For a more familiar example, think of a political election. Could you imagine some news channel calling the election after 0.00013793% of the votes were counted and do so with 95% certainty?. It doesn't happen. It doesn't because the assumptions made on a 800 of 5.8 mil sample outweigh the actual data value of the sample.

It is more and more clear that we are arguing the difference between a class room theory and an actuality.

 

[smokey]

First I have little doubt that polls can be skewed by the questions, language, context and even tone they are presented in. Further steps can be easily taken to knowing skew these polls through targeting respondants (geographically, religious associations, or even time of day if handled via telephone implemented polls).. So lets just say there are literally tons of ways to skew the data and agree on that point.

My issue here is that the sample size is no where near the 3.5% percentage indicated by Tyeraxus post nor does 0.00013793% of a population (800 of 5.8mil) guarantee of confidence interval of 95%as implied.

Confidence intervals are not simply proportional to the number respondants or even percentage at this low of a level. For a more familiar example, think of a political election. Could you imagine some news channel calling the election after 0.00013793% of the votes were counted and do so with 95% certainty?. It doesn't happen. It doesn't because the assumptions made on a 800 of 5.8 mil sample outweigh the actual data value of the sample.

It is more and more clear that we are arguing the difference between a class room theory and an actuality.

Why argue whether the poll is valid or not in the first place? The 2A is a civil right. If a poll asked how many people were in favor of removing the right to vote from blacks, how much legal authority would that carry with it?

p.s.- people are still actin the fool in comments. Here's some good ones...

You guys are wannabes. You watch Dirty Harry and think you can defend themselves with a gun. You are the impotent NRA white boys because you live in a fantasy world of Gawd wants me to have a gun, I've gotta defend myself from the big bad government. Youpretend to know the Constitution, they don't and I laugh when you misquote Thom. Jefferson. You gotta love the moron who posted a video that he was going to start killing people if gun laws changed, a real revolutionary. You read these blogs and its like the question and answer session for the Village Idiot contest. 800 is a big enough sample from the genius statisticians, pathetic.

Don't worry gun boys. The NRA, a murderer's best friend will continue to make sure that VA is an easy buy states so that criminals can buy guns for resell in Maryland and up the East Coast. The NRA promotes the proliferation of guns in American because they are a lobby for the gun makers, they do not care about the Constitution nor the carnage. You are not going to defend yourself with a gun, you will not keep the so-called evil Government at bay. You gun boys say that there is no infringement on your gun rights and in the next breath you say there are 80,000 gun laws in US. I hope the NRA continues with their dogma so that this President will achieve greatness when he uses Executive orders to stop your murderous reign over America, the we can have a better chance of "life, liberty and the pursuit of happiness".

 

[Haides]

people are still actin the fool in comments. Here's some good ones...

It upsets me greatly that these people are allowed to breed.

 

[Markp]

I just read this too. What a pile of poo. 800 people is not a large group.

But they surveyed the people going to the JHU summit on firearms violence so they could get a diverse sample.

 

[jrumann59]

HAHA, I wonder if they called anyone outside of PG or MoCO