Gun Control & RKBA
Related: About this forumWhat is your opinion of this article?
Packing heat may backfire. People who carry guns are far likelier to get shot and killed than those who are unarmed, a study of shooting victims in Philadelphia, Pennsylvania, has found.
It would be impractical not to say unethical to randomly assign volunteers to carry a gun or not and see what happens. So Charles Branas's team at the University of Pennsylvania analysed 677 shootings over two-and-a-half years to discover whether victims were carrying at the time, and compared them to other Philly residents of similar age, sex and ethnicity. The team also accounted for other potentially confounding differences, such as the socioeconomic status of their neighbourhood.
Despite the US having the highest rate of firearms-related homicide in the industrialised world, the relationship between gun culture and violence is poorly understood. A recent study found that treating violence like an infectious disease led to a dramatic fall in shootings and killings.
Overall, Branas's study found that people who carried guns were 4.5 times as likely to be shot and 4.2 times as likely to get killed compared with unarmed citizens. When the team looked at shootings in which victims had a chance to defend themselves, their odds of getting shot were even higher.
http://www.newscientist.com/article/dn17922-carrying-a-gun-increases-risk-of-getting-shot-and-killed.html
ProgressiveProfessor
(22,144 posts)ZombieHorde
(29,047 posts)ProgressiveProfessor
(22,144 posts)discntnt_irny_srcsm
(18,480 posts)...It looks to me like it'll get hits, sells papers, make money...
Might be total hooey but that might crimp the mirth.
CreekDog
(46,192 posts)next!
discntnt_irny_srcsm
(18,480 posts)...WHAT I THOUGHT A YEAR AGO.
gejohnston
(17,502 posts)it isn't the number of guns, it is who has the guns. I'm going out on a limb here and guess these people were not licensed CCW, target shooters, nor country folks who hunt and shoot at beer cans. While they might have been shot while carrying, carrying didn't get them shot. Most likely it was the business they were in, or something else that put them in the triangle of stupid, which increased their chances of getting harmed carrying or not.
If they are trying to say that taking my pistol to the range increases my chances of being harmed even though I don't do anything different than when I don't have it, then I call bullshit.
Now if they are saying that people who spend a lot of time in the triangle of stupid or involved ins some illicit business are likely to arm themselves, and still get shot anyway, then it is obvious. For a publish or perish career, providing scientific evidence of the obvious fills a square.
slackmaster
(60,567 posts)rl6214
(8,142 posts)GreenStormCloud
(12,072 posts)Guns are not bullet magnets. Having a gun on you is not going to make bullets turn in mid-flight and home in on you. The difference is in the behavior of the armed people. If you will look at the details of that study you will find that most of the armed shooting victims are criminals. Being a criminal is a very high-risk occupation. Drug dealers settle business disputes with guns. Gang members go hunting for each other. It isn't the gun that gets them shot, but what they are doing while they have the gun.
I carry a gun all the time, but I am not involved in gangs, drugs, or other criminal activity. I don't hang out with rough crowds. So my probability of being a target is limited to being a victim of random crime. Nobody is going to be hunting for me.
Notice that the study does not have an entry of CCWers.
Here is a link for the study: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2759797/
From the study:
However, compared with control participants, shooting case participants were significantly more often Hispanic, more frequently working in high-risk occupations1,2, less educated, and had a greater frequency of prior arrest. At the time of shooting, case participants were also significantly more often involved with alcohol and drugs, outdoors, and closer to areas where more Blacks, Hispanics, and unemployed individuals resided. Case participants were also more likely to be located in areas with less income and more illicit drug trafficking.
The medical types that did the study discovered what any criminologist learns in Criminology 101. Being a criminal is very dangerous.
DanTex
(20,709 posts)They controlled for various confounding factors, including all of those things that were listed in that paragraph you quoted (high-risk professions, prior arrests, education levels, drug and alcohol use, etc.), and found that even after controlling, the possession
We've been over this before. The problem is, you don't understand how multivariate regressions work, and rather than try and learn, you would rather use your own ignorance as an excuse to dismiss scientific evidence that contradicts your political beliefs.
When we only considered independent variables that most strongly affected our models, smaller but correspondingly significant adjusted odds ratios were noted. In these reduced models, individuals who were in possession of a gun were 2.55 (95% CI = 1.00, 6.58) times more likely to be shot in an assault than those not in possession. Individuals who were in possession of a gun were also 3.54 (95% CI = 1.18, 10.58) times more likely to be fatally shot in an assault. In assaults where the victim had at least some chance to resist, individuals who were in possession of a gun were 2.92 (95% CI = 1.01, 8.42) times more likely to be shot (Table 2 ).
GreenStormCloud
(12,072 posts)Guns are not bullet magnets.
DanTex
(20,709 posts)The part I highlighted says that even after controlling for confounding factors, the correlation still holds and is still statistically significant. So, yes, it does distinguish between legal and criminals. Or at least between people who have a criminal record and people who don't.
GreenStormCloud
(12,072 posts)DanTex
(20,709 posts)Table 1 lists all the variables that were taken into account, including prior arrests, alcohol involvement, illicit drug involvement etc.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2759797/table/tbl1/
petronius
(26,602 posts)The study does not control for time and place. The authors invoke stray bullets to argue that residents of Philadelphia, Pennsylvania, are at equal risk for being shot, no matter where they are and what they are doing. This ignores the fact that violence is not randomly distributed and is unfair to Philadelphia.
The control group is inappropriate, as was probably guaranteed by its selection from all adult Philadelphians. There were large differences between case participants and control participants in prior criminal history and alcohol or drug involvement, all of which influence gun-carrying behavior and risk for violent victimization. Personal and geographic differences compounded one another: 83% of shootings occurred outdoors, yet while those shootings were occuring, 91% of control participants, arguably at lower risk already for personal reasons, were indoors. A list could easily be made of likely differences between case participants and control participants that were not addressed.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2866589/
But what I liked about the original article was some of the concerns they raised themselves:
Situations in which the victim had at least some chance to resist may have generated gun assault risks when one considers that many of these events were 2-sided situations in which both parties were ready and mutually willing to fight on the basis of a prior argument.29,30 Because both victim and offender had some sense of each other's capabilities prior to the event they may have had more time to prepare for their ensuing conflict.61 More preparation may have increased the likelihood that both individuals were armed with guns and that at least 1 or both were shot.
Reinforces the message - that every gun owner should know - that a firearm is just part of a defensive strategy and a substitute for awareness and avoidance...
DanTex
(20,709 posts)For example, studies have found also that keeping a gun in your home increases your chances of being a homicide victim.
The methodology in this study is case-control, which means they look at people who got shot (cases), and compare them to people who didn't get shot (controls) to see what differences there are between the two groups. This is the same kind of study that was originally used to find the link between smoking and lung cancer.
There is always the possibility of confounding or reverse causality issues in these kinds of case-control studies. For example, it may just be that people are at greater risk for homicide to begin with are more likely to carry a gun, and not that carrying a gun makes a person at greater risk of homicide. However, the people doing the study know about this problem, and so they control for such possibilities statistically. For example, the factors controlled for in this study include things like drug and alcohol use, criminal history, being indoors versus outdoors, having a high-risk-profession, attributes of the neighborhood, etc. In other words, the results of this study aren't simply due to the fact that people with a criminal record are more likely both to get shot and also to carry a gun.
Of course, the fact that the study is controlled doesn't mean that the correlation they found is necessarily a causal one. There could still be reverse causation, or there could be other confounding factors they didn't take into account. In fact, even in theory it is not possible to establish causal relationships from purely observational studies. For that, you need to do something like a randomized clinical trial, where you take a group of subjects, give half of them guns and tell them to walk around in the streets and see which group gets shot more often. But that is out of the question for both practical and ethical reasons.
So, while this study alone is not conclusive, it is consistent with other studies relating guns and personal safety. In fact, there are zero studies (that I am aware of) that actually find that owning a gun provides a safety benefit. There are obviously some instances of successful self-defense with a gun, but the studies I've seen that measure the net safety impact of a gun have generally found that a gun reduces safety, rather than improves it.
GreenStormCloud
(12,072 posts)Actually it is homicide victims have a greater chance of having a gun in the home. Since over 50% of homicide victims are themselves criminals, it is reasonable that such people would have guns.
Domestice violence households often have guns. That does not meant that the gun causes domestic violence, but rather that violent people will usually have guns.
DanTex
(20,709 posts)If A is correlated with B, then B is correlated with A. I would suggest familiarizing yourself with basic statistical facts before trying to "debunk" statistical studies that you disagree with politically.
And by the way, as I've pointed out before the studies finding that gun in the home increases your change of being a homicide victim are controlled for things like criminal history.
GreenStormCloud
(12,072 posts)It is indeed dangerous to live with a violent criminal, especially if he has a gun. Law-abiding people rarely misuse their guns. You try to pretend there is no difference between the two sets.
DanTex
(20,709 posts)I'm not sure why you do that. I think I've said it already four times this thread, and you keep ignoring it. So here goes again: the studies are controlled for criminal records. Do you really not understand?
gejohnston
(17,502 posts)how was it controlled and how well controlled?
DanTex
(20,709 posts)of gun ownership, and the effect of criminal history on homicide risk (and some other factors), as well any correlation between gun ownership and criminal history (and the other factors), all at the same time.
So that means that, if the only reason a correlation shows up between gun ownership and homicide victimization is due to the fact that people with criminal records are more likely to own guns and also more likely to be homicide victims, the multivariate model would figure this out, and the correlation between gun ownership and homicide victimization would disappear in the adjusted odds ratio.
This sort of multivariate analysis is done in just about every observational study, and it is necessary, because there are almost always some potential confounding factors.
Now, nobody is claiming that these case-control studies definitively prove a causal link between owning a gun and being a homicide victim. It is possible, for example, that there are other factors that weren't controlled for. But you can't just dismiss the results so easily by saying that the reason the correlation shows up is because criminals are more likely to own guns and also more likely to get shot. And then there is the fact that there are now several studies with similar findings, but no studies (that I know of) have found any safety benefit from a gun.
For the Kellermann study, there is further evidence that the study was properly controlled. The additional homicide risk was all intimate homicide. If the study were not properly controlled, and it were really true that the correlation between guns and homicide was simply a side-effect of the fact that criminals are more likely to own guns than non-criminals, they you'd find that the risk of all different kinds of homicide go up with gun ownership, not just intimate homicide by gun.
GreenStormCloud
(12,072 posts)You ignore the obvious. Guns aren't bullet magnets. Bullets don't curve in the air and go to gun owners. Being a violent criminal is dangerous to both the criminal and those around him/her. The table you posted was a breakdown of percentages, not a claim of all the confounding factors that were considered.
DanTex
(20,709 posts)As I keep repeating, the studies control for criminal records. Nobody is pretending there is no difference.
And yes, the factors listed in that table are the confounding factors used to build the multivariate model. They present the raw percentages in that table as part of the write-up, but then they use the data to do the multivariate analysis. When the study says "we adjusted for potential confounding factors", what do you think that means?
GreenStormCloud
(12,072 posts)The table didn't do that. Further, I expect to see the math .
Carrying a gun in my pocket doesn't make bullets seek me out. I have to do something to make a thug want to shoot me. It is those actions that cause folks to get shot, not the mere possession of a gun.
For most of the people shot, the action they did was being part of the criminal underworld. Violent criminals do violent criminal things and they carry guns - illegally.
Show me a study that tracks only people with CCWs in states with shall-issue permits that require the FBI background check and classes to get the permit. The Joyce Foundation won't fund such a study because they already know the answer it will yield.
DanTex
(20,709 posts)And do you really think they got the math wrong? The stuff they are doing is not some fancy cutting-edge statistics. The calculations are done using some statistical software. It is not their job to explain basic statistics to people who don't understand the first thing about either regressions or case-control studies.
And yes, I get that you aren't going accept the results of any studies that don't confirm your ideological biases.
GreenStormCloud
(12,072 posts)Why does a person with a gun have a 4.5 times greater chance of getting shot? Without an explanation of how it happen, I will dismiss it.
What I see is that violent criminals have a much greater chance of getting shot, and because of the violence of their profession they are far more likely to be armed.
They included people who, by mutual agreement, shot at each other. Law abiding people don't do that. That is a description of two criminals who were out for each other. Since both are packing, naturally somebody with a gun is going to get shot. That's kind of obvious, don't you think? Violent criminals shooting each other - who would have thought they would do that?
Show me a study of ONLY CCWers over a period of five years. That should not be hard to do. There are between 8 to 10 million of us. As long as you include criminals in your study I will continue to disregard the study.
Criminology is a well studied field. I will prefer to listen to criminologists talk about violence than to an epidemiologist trying to play criminologist.
DanTex
(20,709 posts)You don't really care about statistical techniques or controlling for confounding factors or any of that. I'm sure you are aware that this is hardly the only study on gun violence, but like many of the pro-gun extremists, no amount of evidence is going to sway your views. You are going to listen to the NRA and not the scientific community, period.
GreenStormCloud
(12,072 posts)I see no evidence that my gun casts a magic shoot-me field.
You are trying to claim that having a gun causes a person to get shot.
Response to DanTex (Reply #11)
Name removed Message auto-removed
rrneck
(17,671 posts)discntnt_irny_srcsm
(18,480 posts)You'd go there early packing heat and kill everyone who showed up.
<= but was it really needed?
rrneck
(17,671 posts)More likely to scatter marbles on the floor and watch them fall on their ass.
discntnt_irny_srcsm
(18,480 posts)...double secret probation.
Response to ZombieHorde (Original post)
Lizzie Poppet This message was self-deleted by its author.
Eleanors38
(18,318 posts)what are the data on their deaths by firearms, or any other factor, when compared with the general population? Has a study like this been done?
spin
(17,493 posts)of those who have a Texas carry permit with the general population of Texas.
It's quite interesting and you will never hear it mentioned by the national media probably becasue it doesn't paint people with a carry permit in a bad light.
http://www.txdps.state.tx.us/rsd/chl/reports/convrates.htm
Starboard Tack
(11,181 posts)The problem here is not how many people have CC permits, nor is the problem not getting shot. The problem is people being shot and the study shows a 450% increase in likelihood that the person shot was carrying a gun.
That is the fact that needs to be addressed here.
There is no suggestion that a specific individual's chances of being shot are increased by carrying. But we can draw the conclusion that the risk is increased by that factor, across the board.
I think it is due to multiple factors, including, but not limited to the following:
1. Carrying creates a false sense of security and often a misplaced sense of bravado.
2. The idea that carrying a gun somehow equalizes the odds when encountering predators.
3. The idea that carrying a gun will reduce one's vulnerability, rather than increase it.
4. The notion that one is actually prepared to shoot another human, before the other human shoots them.
5. The belief that carrying will serve as a problem solver, rather than a problem creator or enhancer.
6. The belief that carrying a gun means that gun can never be used against its owner.
7. The belief that carrying a gun is more likely to resolve a bad situation, rather than escalate it.
Again, I'm not suggesting that any individual here believes any of the above, but many who carry do. And some who believe may be justified in their belief. The numbers refute that, of course. It would appear that a whole bunch are bullshitting themselves.
My advice to anyone contemplating carrying, would be to consider those 7 factors.
By all means, acquire a CC permit, you never know when you might need it, but think at least 7 times before using it.
gejohnston
(17,502 posts)Last edited Thu Jan 23, 2014, 05:49 PM - Edit history (2)
by an ER doc named Author Kellerman using CDC funds. that was the "43 more times" study. Although it was published in a medical journal, it attracted a lot of attention criminologists. Although it might have received rave reviews among the readers of NEJM, who didn't understand what they were reading, and gun control advocates. Criminologists on the other hand, savaged it. When criminologists asked him for the raw data, he told them to pound sand. In 1993, he "confirmed his own study" and published it in NEJM, but this time "43 times more likely" became "2.7 times more likely".
It was (along with the CDC funded studies in general), in the words of criminologist James Wright during congressional hearings on CDC funding, "it was as scientific as popular literature produced by the NRA." He wasn't the only one that found the study an invalid croc of shit. How did it pass peer review? Because it wasn't because the raw data that Kellerman used wasn't released until 1997.
Part of a counter or critique written by Henry E. Schaffer, Ph. D in the Dec 1993 NEJM (based on what he read in the article)
The Kellerman, et al (1993) study in the NEJM didn't use the same calculation that is shown above. They used the "Mantel-Haenszel chi-square analysis for matched pairs" but didn't give any analysis. This analysis is able to adjust for differences in stratified data *if* the stratification (subdivision of the overall population into the two subgroups) is known and is taken into account when matching..
Matching control pairs is an attempt to get the each case and matched control be in the same subgroup - when the population is divided into subgroups. If this is done, then it appears that the Mantel-Haenszel analysis will produce an association calculation which is free of the confounding demonstrated above. However it is not clear that the Kellerman, et al matching does select controls from the same subgroups as the cases. The control selection was done using a random selection starting outside a "one-block avoidance zone" away from the case homicide, and the matching criteria did not include any life-style or related indicators.
If the population is composed of subgroups which differ in homicide rates, then the matching procedure would be hoped to select the matching control from the same subgroup as the case it is supposed to match. This could happen with the matching method used if the subgroups were settled in distinct different large geographic areas. Because of the avoidance method used these areas would have to be larger than one-block in size (how much larger is hard to tell, since the paper doesn't say how far outside the zone it was necessary to travel to find a matching control who would agree to cooperate.) But it doesn't appear that risk sub-groups are distributed in such a coarse-grain manner. I discussed this with a colleague who is a sociologist/ criminologist who pointed out that risk subgroup factors (drug dealing, violent criminal events, violently abusive family relationships, etc.) often are fine-grained. They vary between different families in one apartment building, and certainly vary between different families in a block. Therefore choosing a control who lives 1 or more blocks away will not assure matching with respect to the sub-group. For a minority sub-group (e.g. the 10% "High Risk" group in the examples above) the chances good are that homicides in the high risk group will be matched with low risk group controls.
The Kellerman, et al paper presented all of its data in terms of the overall group numbers, similar to the total population information presented in the examples above. Therefore there is no way to rework the analyses and check on the Mantel-Haenszel analysis results. Without proper within-sub-group matching the Mantel-Haenszel result would be affected by confounding and therefore produce incorrect results just as found by the odds ratio analysis used in the above examples.
That was the $27M study that took away the CDC's lobbying funds.
2. The idea that carrying a gun somehow equalizes the odds when encountering predators.
3. The idea that carrying a gun will reduce one's vulnerability, rather than increase it.
4. The notion that one is actually prepared to shoot another human, before the other human shoots them.
5. The belief that carrying will serve as a problem solver, rather than a problem creator or enhancer.
6. The belief that carrying a gun means that gun can never be used against its owner.
7. The belief that carrying a gun is more likely to resolve a bad situation, rather than escalate it.
Again, I'm not suggesting that any individual here believes any of the above, but many who carry do. And some who believe may be justified in their belief. The numbers refute that, of course. It would appear that a whole bunch are bullshitting themselves.
hack89
(39,171 posts)Andrew Papachristos, an associate professor of sociology at Yale, analyzed police and gun homicide records from 2006 to 2011 for people living in a high-crime neighborhood in Chicago. He found that 41 percent of all gun homicides occurred within a network of less than 4 percent of the neighborhood's population, and that the closer one is connected to a homicide victim, the greater that person's chances were for becoming a victim. Each social tie removed from a homicide victim decreased a person's odds of becoming a victim by 57 percent.
"What the findings essentially tell you is that the people who are most at risk of becoming a victim are sort of surrounded by victims within a few handshakes," Papachristos says. "These are young men who are actively engaged in the behaviors that got them in this network."
Papachristos says it makes sense to look at the spread of gun violence like the spread of a disease or an epidemic, comparing it to how people contract HIV. Much like the roles needle sharing and unprotected sex play in the spread of HIV, a person's behaviors and personal associations play a role in the spread of gun violence homicides, he says. "It's the behavior of sharing needles that puts you at risk for contracting HIV, not simply being poor and black and living in a certain neighborhood," Papachristos says. "The same is true with violence. It's who you hang around with that gets you in trouble."
http://www.usnews.com/news/articles/2013/11/14/gun-violence-significantly-increased-by-social-interactions
interesting study that perhaps points out a way to predict and possibly stop gun violence.
The Chicago police actually used the study to target gang members:
The effort by Austin District Cmdr. Barbara West comes as Chicago police try to quell gang violence with both innovative ideas and more traditional methods such as paying hundreds of officers to work overtime on their days off.
The strategy is based on research by Andrew Papachristos, a sociology professor at Yale University who conducted a study on the West Side several years ago that showed that much of the violence involved a relatively small number of victims and offenders.
A computer analysis showed a little more than 400 people across the city including about 25 in the Austin District are the most at-risk from among more than 16,000 gang members. There can be many reasons for making the list, including lengthy arrest records, the serious nature of the offenses and whether they have been shooting victims themselves.
http://www.chicagotribune.com/news/local/ct-met-chicago-police-heat-list-20130721,0,3279169.story
Starboard Tack
(11,181 posts)They tend to support the old adage, "If you play with fire, you're more likely to get burnt, sooner or later". That applies to both studies, of course.
Ironically, this adage also applies to fire extinguishers. I carry several on board my boat and keep a couple real handy. I had to use one a couple of years ago when a fuse melted and caused an electrical fire. Worked great, but what a mess to clean up. On another occasion, in a rough sea, I had one break loose from its bracket and self-activate. Took a week to clean up that mess.
I guess what I'm trying to say is, that too much "safety" can be dangerous, like too much of anything else.
Eleanors38
(18,318 posts)that some how folks here, as you suggest, missed out on the misconceptions you list; that hardly seems likely if your notions are general truths.
Respectfully, the study goes to a rather small group of people living, it seems clear to me, in an intensely criminal environment. Be in or near that environment and your chances if being shot go up; remove yourself from it, your chances go down. Some of your listed assertions may indeed apply to those within that concentrated milieu. But they apply to few of the millions of legal concealed-carry folk.
Bearing arms legally is an eminently personal act, based on individual assessments of dangers. The truth of the listed assertions (6 - 7) would have to be borne out within the community of Legal concealed-carry folks. It hasn't. As to (4), whether self-defense experts recommend it or not, many threats are thwarted by display of a gun or even announcing you have one; (5) misses the point: possessing a gun for SD is not social policy. Whatever problems are created will take a backseat to the primary purpose of self-preservation.
Starboard Tack
(11,181 posts)You say it is based on "individual assessment of danger". Are you inferring that those who carry routinely, wherever they go, have assessed the entire world as being so dangerous, that they need a gun to deal with it on a daily basis? That's quite an indictment of the society you live in.
How would you apply this dictum to the "road rage" incident in Houston? Should the victim have responded in kind?
Imagine a world where everyone who is legally qualified is carrying. Is that really what you want?
Eleanors38
(18,318 posts)If someone had a gun in my face and I couldn't defend myself, I would try to defuse the situation and select a 180 caliber.
I haven't thought much about everyone legal carrying. It seems to bother you. Why is that?
Starboard Tack
(11,181 posts)And my question was not about being unarmed. That's an easy answer and as we saw, IRL it worked out OK, probably because the victim was unarmed, thus unable to easily escalate the situation.
My hypothetical question about everyone carrying was asked in the hope that you might share your thoughts on the subject. I didn't assume that you had already pondered the question. But I hoped you might give the question some serious consideration. I'm trying to get my head around the mindset of habitual carry and the potential impact it might have on society. This is an area where you and I obviously disagree, but I am open to being persuaded differently, if I see the logic, which , so far, I don't.
Eleanors38
(18,318 posts)constructive and serious about this) is solicit someone from GD (or elsewhere) and ask that person why. I would also suggest finding other terms than "mindset" when inquiring. Good luck.
Starboard Tack
(11,181 posts)Well, if anyone here who does carry on a regular basis, and I know there are quite a few, they can jump in and answer my question. Don't see much point of taking this to GD, where the overwhelming majority don't carry.
I don't find the thought of carrying as outrageous as most folk and often imagine myself in situations like this guy found himself. That's why I ask the question "What would I do?" Especially with my terrified wife and child in the line of fire.
sarisataka
(18,679 posts)and I'm not sure what is your specific question. It sounds like you might have a good point deserving of an answer. Might I suggest putting it up as an OP? I, for one, will give you an answer.
sarisataka
(18,679 posts)the conclusions drawn from it...
Kleck samples 5000 and determines ~2 million DGU - a ridiculous number with a bias due to a small sample size that only supports a political viewpoint.
U Pa samples 677 and determines carrying a gun makes you >4 times more likely to get shot - an unquestionably accurate number that proves what one side has said all along...
DanTex
(20,709 posts)The objections to the Kleck study aren't based on sample size, they are based on the susceptibility of phone surveys measuring rare events to false positives. For example, according to the Kleck methodology, more people are abducted by aliens every years than there are DGUs. It doesn't matter how big the sample is, the fact of the matter is a small percentage of the population will answer "yes."
gejohnston
(17,502 posts)a not award winning criminologist. Oh wait, he got a plaque and a free dinner from the Brady Campaign. That was only Hemenway's objection.
sarisataka
(18,679 posts)it sounds like it was only done in Philadelphia, does the trend hold for other cities, what about rural areas... It seems like there are racial aspects, do those need to be addressed.
I will freely admit that my statistics class was very long ago and I didn't pay much attention then so I can not question the methodology except from a layman's perspective. The conclusion is interesting, but intuitively it seems off. An explanation is needed that can be understood at the layman level. WHY do carriers get shot so much more often? Is it the result of their action or the action of another? What are the demographics and is there a high spike among any socio-economic or other group?
Like so many studies, it creates more questions than it answers yet is assumed it is the final answer by whichever group the conclusion supports.
*for the record I think Kleck's DGU estimate is too high as well.
DanTex
(20,709 posts)The point of my last post is that there are many more issues than just sample size. Particularly since case the Kleck study and this study are truly apples and oranges. The Kleck study was a phone survey. This one was a case-control study.
What that means is that they look at all people who were shot during a certain time period (these are the cases), and then for each case they find someone who didn't get shot (the controls), but in other ways resembled the case person (in this study, it looks like they matched according to things like where the people were walking, age, sex, etc). Since the case-control matching isn't going to account for everything, they also collect a whole bunch of auxiliary variables (things like drug use, criminal history, socioeconomic variables, etc.), and then build a statistical model that tries to isolate the effect of gun possession on risk of getting shot.
A study like this cannot answer the question of why carriers get shot more often. There are some plausible explanations, for example, maybe carrying a gun makes a person less likely to flee and more likely to fight against an attacker.
As far as demographics and socieconomic things, they did look at those variables and controlled for them. But there are some things that are hard to control for. In general, the biggest sticking point in a study like this is not sample size, but whether it is properly controlled.
Similarly, the problem with the Kleck study isn't sample size. The sample was large enough to say with confidence that about 1% of the population (about 2M people at the time) will say "yes" if you call them up and ask them if they've used a gun in self-defense in the last year. The real question is whether the "yes" answers represent legitimate DGUs or not, and there is very compelling evidence that most of them are not.
The case-control methodology is not the best kind of epidemiological design. The "best" kind of study would be a randomized clinical trial, which in this case would mean getting a group of subjects, splitting them at random into a gun group and a no-gun group, and have them all go about their daily lives and see who gets shot more often. But, for a lot of reasons, this isn't feasible.
What's more noteworthy than just the result of this one study is the fact that every case-control study done to examine risks versus benefits of owning a gun has either found a significant increase in risk or inconclusive. There are no studies that have found that carrying or owning a gun makes people safer, at least not that I'm aware of.
sarisataka
(18,679 posts)it makes sense that it would be more accurate- if the control group is valid.
I'll give U Pa credit and assume they know how to do these. The next question then is what is the base number. If the cotrol group gets shot 10% of the time and the increase is 4x+, then we are talking 50% rate. If the control rate is .01% then the odds of being shot go from 1/10,000 to 1/2500- an increase but still rare event.
I know FBI stats say armed resistance is better than unarmed or compliance but don't know the criteria to match against this study.
I'll agree it is an interesting finding and cause for more research. I don't think it is the be all end all statement though we don't have to remain static until all studies are done. Some laws proposed, e.g. UBC, accountability... are a good start and have obvious benefits even without all questions answered.
DanTex
(20,709 posts)You're probably thinking of the National Crime Victimization Survey, which does ask about resistance. The problem with drawing conclusions from NCVS about the efficacy of resistance, as many people have pointed out, is that it is completely uncontrolled and doesn't take into sequencing of events. The raw NCVS data, last I read, without any controlling, indicates that armed resistance may prevent loss of property but doesn't do any better in terms of preventing injury, when compared to unarmed resistance.
But you can imagine the causality issues there. For example, suppose someone would like to resist, but injured first and doesn't have a chance to resist. That would go down as a "injury with no resistance." And even if people aren't injured first, the severity of the attack could affect the decision of whether to resist or not. This could work both ways -- some people might only resist against serious threats, other people might choose not to resist more serious threats for fear of escalation. Etc.
sarisataka
(18,679 posts)I had been thinking NCVS.
Agreed, the lack of control does limit the usefulness of making sweeping generalizations. Also the criticisms that the pro-carry side makes, that NCVS under counts DGU, would apply as well. The reluctance to report could skew the data in either direction. As you mention, there also the issue of property crime vs violent crime. While I may not feel 100% sympathetic to a criminal shot while trying to break into a garage, neither would I believe it was a 'valid' DGU if there was no threat of bodily harm to the shooter.
Good point about level of resistance and level of threat. The case is closed so I can relate my own 'DGU' from last year. I believe it is an example of what you are alluding. I am going to a party for my little BiL who just came home from Boot Camp, but I will try and post the incident later this evening.
petronius
(26,602 posts)people were walking" ; the control group was all adult Philadelphians, and thus excluded neighborhood variability as well as whether the control was indoors or out at the time. If you haven't seen it, the commentary by Garen Wintemute on this paper is pretty informative (I linked it up above)...
DanTex
(20,709 posts)The case-control matching was based only on date, time, age, gender, and race. Things like indoors/outdoors and location ere controlled for statistically, but were not part of the case-control matching.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2759797/
LAGC
(5,330 posts)That is quite possible. I know its anecdotal, but we all saw just recently how cocky Zimmerman got, chasing after Trayvon that dark night. It's pretty doubtful he would have gone traipsing around after a suspected burglar if he wasn't carrying and knew he had that handgun as a backup in case things headed south, which of course they did.
But then you have to weigh in the many cases of gang-bangers. These are folks who tend to pack heat and get in many confrontations, so naturally if they are being included in these studies, there is going to be higher correlation between packing heat and getting shot.
Just food for thought.
DanTex
(20,709 posts)People who are career criminals are more likely to carry a gun, and also to get shot. This is clear. But studies like this one account for the statistically, by controlling for things such as criminal history, drug use, etc.
Of course, there's always still the possibility that the study wasn't controlled enough anyway -- like I said, this is the sticking point with studies like these. For example, even though two individuals have the same arrest or drug record on paper, that doesn't mean both of them are violent gang members, so even after controlling for drugs and arrests there could still be some of the confounding effect you described.
In some studies I've seen, there are indications that this effect was properly controlled for. For example, in a study about guns in a home, they found that the risk of homicide increased, but specifically of homicide by gun committed by another household member. In this situation, if the statistical finding was due to the fact that violent gang members are more likely both to own a gun and also get killed, one would expect that the risk of homicide would increase across the board. Gang member's aren't just more likely to be killed by other family members, the are more likely to be killed by other people as well. So the fact that the increase was only in a specific area is evidence that the controlling worked properly.
For the Philly study, I'm not aware if there is any similar verification of the controlling.