Summations Only

So, you are saying that those who have guns are bad guys and those who don't have guns are good guys, and we should arm the good guys which would then make them bad guys? Seriously, I think I know what you are saying, and at one level, I get it. But here's the problem I have with it. Good, honest citizens DO have guns as well as the bad guys. The difference is that the bad guys are using them to kill people. So what we are wanting is for the good guys to start USING their guns to kill bad guys. Now, I know that was a fun game to play as kids using toy guns, but as adults, using real guns, it is no longer a fun little game.The problem I think is this scenario blurs the distinction between good and bad. I used the Old West as an example, and think about the heroes from those stories. Was Billy-the-Kid a good guy or a bad guy? He killed bad guys, right? What about Wyatt Earp? Doc Holiday? Etc...My son, who is 7, was playing a game on his tablet and me being the parent, has to monitor what he does. He was playing an assassination game where you would use a high powered rifle to shoot people. I told him, that he was not supposed to play games like that and for him to take it off his tablet. He said that it was OK because he was only killing bad guys.Is that what we want? Do we have the hope that eventually all the bad guys will be dead and only the good guys will be left? I just don't think it's the answer.HBD

I have trouble believing anyone would go to such trouble to make up stuff like that, let alone convince themselves it's a good thing to do.It would be nice if the gun cause could be based on statistics but if it can't it can't.

I took the data from the table in the Wikipedia article on Gun Violence in the United States by State and did a scatter plot. There's no apparent relationship between the prevalence of guns and gun murders: Checking to see if there was some other relationship I looked at the top and bottom 10 states for gun murder rates, and I included the largest city in each state. Here are the bottom 10 states and their largest cities:

• Vermont (Burlington)
• New Hampshire (Manchester)
• Hawaii (Honolulu)
• North Dakota (Fargo)
• Iowa (Des Moines)
• Idaho (Boise)
• Maine (Portland)
• Utah (Salt Lake City)
• Wyoming (Cheyenne)
• Oregon (Portland)

And here are the top 10 states and their largest cities:

• Georgia (Atlanta)
• Florida (Jacksonville)
• Mississippi (Jackson)
• Delaware (Wilmington)
• Michigan (Detroit)
• South Carolina (Columbia)
• Maryland (Baltimore)
• Missouri (Kansas City)
• Louisiana (New Orleans)
• District of Columbia (itself)

There's nothing to distinguish the top 10 states from the bottom 10. Each list has some that tend toward the more rural and some that tend toward the more urban. It's only when you look at the largest cities that you notice a difference. States with large cities with histories of urban blight and unrest have the highest gun murder rates. We need to compare like with like, so I need to find statistics by city. Wikipedia has a table that includes murder rates by city in their article List of United States Cities by Crime Rate, but it doesn't include gun prevalence. If anyone knows of such a study let me know.I did find an interesting study (The Relationship Between Gun Ownership and Firearm Homicide Rates in the United States, 1981—2010) that appeared in the American Journal of Public Health. It found that "for each percentage point increase in gun ownership, the firearm homicide rate increased by 0.9%."It's exactly what you would expect, an approximately one-for-one relationship. Increase gun ownership by 1% and you increase the gun homicide rate by around 1%.One of the reasons for the paucity of good studies is the NRA's stranglehold on Congress. If the NRA were really so certain that statistical studies would confirm that guns make us safer they wouldn't oppose them.--Percy

Really? I explained the problems with the graph, but you're just going to bury your head in the sand and ignore the evidence of fabrication? Tell you what, why don't you explain that graph to us. Here it is for everyone's viewing pleasure (from Guns in Other Countries): Explain the 3 bars above Saint Kitts and Nevis. Explain why there are 3 bars in no-man's land between England/Wales and El Salvador. Explain why so few country names line up with a bar. Explain why there are 62 countries but nearly 200 data bars.I checked a few of the values for the average firearms per 100 people, and that seems accurate since it corresponds to the values in the Wikipedia article on Number of Guns Per Capita by Country, but the homicide by firearm data is completely bogus.AbE: Then what data are you using to support your conclusions?--PercyEdited by Percy, : AbE.

That's probably false. That may be true. But it isn't true because guns make you safer. It's true because Canadians don't own guns for protection. As I've mentioned before, in my lifetime I've met only one civilian (other than collectors) who owned a handgun. (And he had it for sharks, not people.)

As typical it appears to not be totally bogus, but at best a half-truth, or more likely a mis-representation of the truth. Obviously, who ever made that graph knows nothing about statistics and as you pointed out, did a poor job of laying it out.It appears the data came from Here.There are 186 countries listed but 79 of those countries have a null value for one or both of the data points shown on the chart, so it is inappropriate to include those countries in the chart. For example, Saint Kitts and Nevis has a gun homicide rate of 32.44/100,000 but a null value in # guns/100. Conversely, Benin has a gun ownership rate of 1.4/100 but a null value in homicide rate, which the chart represents as zero.(sorry I don't know how to post this Excel data, so I will just have to describe it)After deleting the data with null values, we are left with 107 countries. A regression analysis shows a slope of -0.085 with a 95% confidence interval of -0.246 to 0.076, an intercept of 5.72 and a p-value = 0.296. So while there is a negative linear correlation, it is not significant. R²=0.010, which means that the linear model only explains 1% of the variation in the data.This chart doesn't say what Faith thinks it does, but not sure if "fabricated" is the right word. Just a terrible mis-representation of the data.HBD(notes for anyone who may not understand the statistical data above)*negative slope means gun homicides decrease as gun ownership increases
** an intercept of 5.72 is the hypothetical value at which the regression line crosses the 0% gun ownership value axis.
*** for the linear relationship to be significant the p-value needs to be less than 0.050 for 95% confidence.
**** the confidence interval means that we can be 95% confident that the actual value of the regression slope falls somewhere within that interval. since the interval includes zero and positive values, we can not say for sure that there IS a negative correlation.ABE: even if -0.085 is the correct regression slope, this means that as gun ownership increases by 10%, gun homicides are reduced by 0.85/100,000. Is that really a worthwhile decrease in gun homicides? Edited by herebedragons, : No reason given.Edited by herebedragons, : Ooops. An R² value of 0.010 means the linear model explains 1% of the variation, not 10% as I had previously wrote.

I noticed that the District of Columbia is by far the highest for gun murders/capita (an extreme outlier for sure). I wonder if there is a correlation between the number of government officials in a state and the number of gun murders HBD

The data in that first chart from Guns in Other Countries does appear to be at this link that you provided: World firearms murders and ownership. How did you find it? The reference from the website for that chart in footnote 4 does not contain that data.The data you found *does* appear to be pretty accurate. I found a few differences with the data from the Wikipedia article on List of Countries by Firearm Related Death Rate, but not enough to be worrying about possible fabrication.There is, as you said, no meaningful relationship, but if instead of plotting the data for all countries you plot the data for countries similar to the United States so that their data is meaningfully comparable you get this scatter plot: I think if you put that through your tool you'll get a pretty good correlation. The countries I chose were:

• Australia
• Austria
• Belgium
• Canada
• Czech Republic
• Denmark
• England and Wales
• Finland
• France
• Germany
• Greece
• Hungary
• Ireland
• Israel
• Italy
• Japan
• Lithuania
• Luxembourg
• Netherlands
• New Zealand
• Northern Ireland
• Norway
• Poland
• Portugal
• Puerto Rico
• Romania
• Scotland
• Slovakia
• Spain
• Sweden
• Switzerland
• Turkey
• United States

--PercyEdited by Percy, : Wordsmithing.

I'm back to simply not trusting statistics. I can't analyze them myself and even from the discussion between you and HBD it appears there are just too many ways to get things wrong whether fabricated or not.I can barely see the writing on that graph let alone even begin to have a way to interpret it. Sorry.Are all the terms even standardized? I mean, if a study is counting "homicides" do they separate real homicides or killing of innocents, from deaths of a criminal in the act for instance? Don't answer, that's just one of too many questions that need answering to make statistics meaningful. But my general question is whether they distinguish between the good guys and the bad guys, you know, my favorite distinction.Somewhere on that page it says that the UK only counts homicides if that's been established in court, but the US counts them at the time of commission. If so that would make for a discrepancy that renders all US-UK comparisons invalid.Is the lack of good statistics REALLY the fault of the NRA somehow? I don't trust statements like that either.I thought I'd found some decent information at FactCheck.org but then found a site that exposed it as a leftist site. That may or may not mean the information is bad but it means I have to do more research than I'm up to, in an area I don't feel I have any ability to judge.Then you ask how I can support my convictions without statistics? Well, I do what I can along these lines:1. It's a Second Amendment right. The buck stops there.2. All the people I know who have possessed and do possess guns have done so safely, taking them for granted as a part of life, and I don't want to see such people treated as criminals by gun control freaks.3. Accidents with guns are tragic but you can't use that against the majority of responsible gun owners. If there is any way to improve gun safety, say by more training or finding ways to keep them out of the hands of irresponsible people without penalizing the good guys, I'm for it.4. Homicides with guns are tragic but you also can't use that against the responsible gun owners, who are the vast majority. The fact that there are bad guys has nothing to do with the good guys who are only interested in valid uses of guns, including self defense and protecting others. Whatever statistics you find on this do not justify penalizing the majority of responsible citizens by the minority of criminals. Already the gun control laws have given criminals access to schools and other locations where they can murder at will because there are no good guys there to stop them, and that's the fault of gun control, not guns.5. I do believe it may yet be possible to put together a valid statistical picture that would support the claims of the gun owners, that we are safer with than without citizen possession of guns. Because there are situations that don't get into the statistics such as how many crimes were prevented by guns, and so on. I've heard many times of this area in the US where gun ownership is very high and crime just about zero and it's frustrating that I can't find information on it now.6. And please do remember that the foundational reason for the Second Amendment was the lesson learned in Europe, and in fact in the Colonies under King George, that if the people don't own means of self-defense it's their own government that will subjugate them. That is not something you can measure by statistics at this time, but the pollyanna insistence that we're silly to think it could happen in our modern civilized world flies in the face of the lessons of history. For all we know it could be the possession of guns by American citizens that has prevented such a situation here already, and it's been claimed by some that foreign takeover would have been attempted by now if it weren't for our citizen guns. There are still bad guys in the world you know, and this is still a fallen world where even seemingly benign leaders can become tyrants, because they think they know better than everybody else how things should be run, and are willing to make us all do it their way by whatever means it takes. The whole point of the American separation of powers is to keep all us potential tyrants from getting the upper hand over all the rest.Edited by Faith, : No reason given.

Amazing tranformation. All it took was showing you that your own statistics were nonsense. You don't need to trust such statements. You should be able to find out the truth yourself. If you are more interested in finding out the truth than in maintaining plausible deniability doing searches for the CDC and Jay Dickey might be enlightening. The second amendment does not prevent enacting gun control measures.The real problem with the rest of your explanation is that there are statistics and they aren't in favor of your position. Those stats also tell us that your anecdotes aren't worthwhile.I appreciate the insight into your thinking. I especially enjoyed the belittling of positions you disagree with. Don't ever ask for respect again.

Rewrite: OK I'm going to apologize for losing my temper at you, NN, for your trashing of what I thought was a perfectly good post explaining how I don't trust statistics but how gun rights can be argued without them. You said nothing substantive that could be answered. I should have ignored you.Edited by Faith, : No reason given.

The link to the data is in The Guardian article listed on the chart. If you're interested, I use MegaStat for Excel for quick and simple data analysis. I got it with a textbook I bought a few years ago, but I think you can get it for free here. MegaStat is not particularly powerful, but it is easy to use. A more powerful statistical software package that I use is R. You can find it here, also free. It is a lot more difficult to use, but is more powerful, accurate and reliable (if you can figure out how to use it ).When I restrict the data set to the countries you listed there is a positive correlation. (slope = 0.022, intercept = -0.038, p < 0.001, R² = 0.552). When I use the "force intercept to zero function", the correlation is even better (slope = 0.021, p < 0.001, R² = 0.706). I am not sure there is an objective reason to restrict the dataset in this way. What is the criteria for being "similar to the US?"There just doesn't appear to be any good correlation between gun ownership and firearm deaths on a worldwide scale. Perhaps if we had a similar dataset of cities within the US?HBD

To get your spreadsheet posted here I think you can copy and paste it into Paint and save it from there to post here. If necessary the Paint image can be posted to something like Photobucket first but I think you might be able just to post it straight with the img code. I keep forgetting the routines I've used. Or if copy and paste doesn't do it, try a screenshot and copy it into Paint. Or if you don't have Paint... I would copy mine into my blog and then to Photobucket but if you don't have a blog...Edited by Faith, : No reason given.

Good to know.

Have you ever been to any of those cities Percy? I've been to most (both the top and the bottom) and the main difference is very easy to spot. It's night and day (or black and white). The problem is cultural and no amount of gun laws will fix that.