Penn Study Asks, Protection or Peril? Gun Possession of Questionable Value in an Assault

[OKIsItJustMe]

http://www.uphs.upenn.edu/news/News_Releases/2009/09/gun-possession-safety/

September 30, 2009

CONTACT: Karen Kreeger
[phone]
[e-mail]

Penn Medicine - University of Pennsylvania School of Medicine and University of Pennsylvania Health System

Penn Study Asks, Protection or Peril? Gun Possession of Questionable Value in an Assault

Those possessing gun in assault situation 4.5 times more likely to be shot than those not possessing one

PHILADELPHIA – In a first-of its-kind study, epidemiologists at the University of Pennsylvania School of Medicine found that, on average, guns did not protect those who possessed them from being shot in an assault. The study estimated that people with a gun were 4.5 times more likely to be shot in an assault than those not possessing a gun.

The study was released online this month in the American Journal of Public Health, in advance of print publication in November 2009.

“This study helps resolve the long-standing debate about whether guns are protective or perilous,” notes study author Charles C. Branas, PhD, Associate Professor of Epidemiology. “Will possessing a firearm always safeguard against harm or will it promote a false sense of security?”

What Penn researchers found was alarming – almost five Philadelphians were shot every day over the course of the study and about 1 of these 5 people died. The research team concluded that, although successful defensive gun uses are possible and do occur each year, the chances of success are low. People should rethink their possession of guns or, at least, understand that regular possession necessitates careful safety countermeasures, write the authors. Suggestions to the contrary, especially for urban residents who may see gun possession as a defense against a dangerous environment should be discussed and thoughtfully reconsidered.

A 2005 National Academy of Science report concluded that we continue to know very little about the impact of gun possession on homicide or the utility of guns for self-defense. Past studies had explored the relationship between homicides and having a gun in the home, purchasing a gun, or owning a gun. These studies, unlike the Penn study, did not address the risk or protection that having a gun might create for a person at the time of a shooting.

Penn researchers investigated the link between being shot in an assault and a person’s possession of a gun at the time of the shooting. As identified by police and medical examiners, they randomly selected 677 cases of Philadelphia residents who were shot in an assault from 2003 to 2006. Six percent of these cases were in possession of a gun (such as in a holster, pocket, waistband, or vehicle) when they were shot.

These shooting cases were matched to Philadelphia residents who acted as the study’s controls. To identify the controls, trained phone canvassers called random Philadelphians soon after a reported shooting and asked about their possession of a gun at the time of the shooting. These random Philadelphians had not been shot and had nothing to do with the shooting. This is the same approach that epidemiologists have historically used to establish links between such things as smoking and lung cancer or drinking and car crashes.

“The US has at least one gun for every adult,” notes Branas. “Learning how to live healthy lives alongside guns will require more studies such as this one. This study should be the beginning of a better investment in gun injury research through various government and private agencies such as the Centers for Disease Control, which in the past have not been legally permitted to fund research ‘designed to affect the passage of specific Federal, State, or local legislation intended to restrict or control the purchase or use of firearms.’”

This study was funded by the National Institutes of Health. The authors are also indebted to numerous dedicated individuals at the Philadelphia Police, Public Health, Fire, and Revenue Departments as well as DataStat Inc, who collaborated on the study.

Therese S. Richmond, PhD, CRNP, School of Nursing; Dennis P. Culhane, PhD, School of Social Policy; Thomas R. Ten Have, PhD, MPH, and Douglas J. Wiebe, PhD, both from the School of Medicine, are co-authors.

###…

http://www.ajph.org/cgi/content/abstract/AJPH.2008.143099v1

[damntexdem]

1. And among the dangers are that the assaulted gun owners will shoot themselves.

Of course, a case-control study does not establish causality; but it is a valued approach to establishing links that can later be tested for causality.

[OKIsItJustMe]

8. From the study

…

RESULTS

Over the study period, our research team was notified of 3485 shootings of all types occurring in Philadelphia. This translated into an average of 4.77 (standard deviation = 2.82) shootings per day, with a maximum of 21 shootings in a single day and an average of 9 days a year that were free from shootings. From among all these shootings, 3202 (91.88%) were assaults, 167 were self-inflicted (4.79%), 60 were unintentional (1.72%), 54 were legal interventions (1.55%), and 2 were of undetermined intent (0.06%). When we considered only assaults, an average of 4.39 (SD = 2.70) individuals were shot per day in Philadelphia with a maximum of 20 in a single day and an average of 13 days a year in which no individuals were shot.

From among all 3202 individuals who had been shot in an assault, we excluded those aged younger than 21 years or of unknown age (29.83%), non-Philadelphia residents (4.34%), individuals not described as being Black or White (1.62%), and police officers that had been shot (0.09%). From the remaining group of 2073 participants, we randomly selected and enrolled 677 individuals (32.66%). We also concurrently identified and enrolled an age-, race-, and gender-matched group of 684 control participants.

Case participants and control participants showed no statistically significant differences in age group, race, and gender distributions, or in the times of day, days of the week, and months of the year when their data were collected. Case participants and control participants were thus successfully matched on age category, race, gender, and time.

…

Association Between Gun Possession and Gun Assault

After we adjusted for confounding factors, individuals who were in possession of a gun were 4.46 (95% confidence interval =1.16, 17.04) times more likely to be shot in an assault than those not in possession. Individuals who were in possession of a gun were also 4.23 (95% CI =1.19, 15.13) 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 5.45 (95% CI =1.01, 29.92) times more likely to be shot.

…

[X_Digger]

14. But not matched on..

-neighborhood crime rate
-previous criminal record
-illegal drug use

So white, male, 35yo junkie who gets shot at a deal gone bad in the hood at 8am is compared to a white, male, 35 year old in the 'burbs.

Makes perfect sense to me.

[OKIsItJustMe]

17. You’re missing a key part of the statement

http://www.ajph.org/cgi/reprint/AJPH.2008.143099v1.pdf

…

After we adjusted for confounding factors, individuals who were in possession of a gun were 4.46 (95% confidence interval =1.16, 17.04) times more likely to be shot in an assault than those not in possession. Individuals who were in possession of a gun were also 4.23 (95% CI =1.19, 15.13) 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 5.45 (95% CI =1.01, 29.92) times more likely to be shot.

…

[X_Digger]

19. Well, since I don't have a subscription..

.. has this been published in a non-subscription journal?

Seems to fall prey to the same foibles as the Kellerman work on the subject.

[OKIsItJustMe]

20. "…has this been published in a non-subscription journal?"

I don't know. I encourage you to search for it. The title is “Investigating the Link Between Gun Possession and Gun Assault.” The authors are Charles C. Branas, PhD, Therese S. Richmond, PhD, CRNP, Dennis P. Culhane, PhD, Thomas R. Ten Have, PhD, MPH, and Douglas J. Wiebe, PhD.

doi:10.2105/AJPH.2008.143099

It is of course possible that the NIH funded a completely invalid study to satisfy the powerful anti-gun lobby, and that the peer-review process failed completely to detect serious problems in the methodology. However…

[X_Digger]

24. My problem with peer reviewed gun studies..

.. is that the peers are usually doctors and epidemiologists- it's like getting a second opinion on your transmission work- from a panel of neurologists.

[OKIsItJustMe]

27. Your objection would be more valid if it was a study of how guns operate

Instead, this is a study of how people operate, when guns are involved.

So, an epidemiologist's opinion is quite appropriate.

[X_Digger]

30. But it's not a disease..

.. peer reviewed by criminologists? That, I can put more faith behind.

[Euromutt]

37. It wouldn't be the first time

It is of course possible that the NIH funded a completely invalid study to satisfy the powerful anti-gun lobby, and that the peer-review process failed completely to detect serious problems in the methodology. However…

It would fact, it would not only not be the first time, I'd be the umpteenth. The public health literature is prone to an unfortunate ideological bias when it comes to research involving firearms. How else to explain the fact that NEJM published studies by John Sloan and Arthur Kellermann in which the lead authors did not deposit their research data with the journal, and refused to make it available to other researchers? What are we to think of a peer reviewer who signs off on a study even though he has no way of telling whether the findings are supported by the data?

Criminological research on firearms produces a diversity of findings, whereas public health research--at least, the stuff that makes it into publication--produces almost uniform findings, readily summed up as Guns Are Bad. Why, it's almost as if everyone's working toward a predetermined conclusion! Perish the thought, right?

A large art of the problem with this particular issue is that it's not a medical one, and medical researchers who delve into it are in effect engaging in social science, not medical science, and many--perhaps most--medical researchers are not scientists to begin with; the know medicine, but they aren't necessarily fully conversant with the scientific method. Which explains much of the argument underlying John Ioannidis' points in his articles "Contradicted and Initially Stronger Effects in Highly Cited Clinical Research" (JAMA, 2005 http://jama.ama-assn.org/cgi/content/abstract/294/2/218) and the more provocatively titled "Why Most Published Research Findings Are False" (PLoS Med., 2005 http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=16060722). And that's when they stick to the stuff they know.

[Euromutt]

46. "We adjusted for confounding factors" -- that's what they *always* say

I'm always highly skeptical that claim in any way legitimizes the findings. From having read quite a few examples of this kind of study (claims that this one is "one-of-a-kind" notwithstanding, because it is no such thing), I rather get the impression that the claim that confounding factors were adjusted or controlled for means little more than "we did some hand-waving and twiddled the numbers until they said what we wanted them to."

Let's see if the results of this study can be replicated, preferably by someone who isn't active in the field of public health. Replicability is the true test of scientific validity.

[PavePusher]

54. "Confounding Factor" = "Something that wouldn't let us get the results we wanted". n/t

[MineralMan]

2. It happens. The mugger bops you on the head, then...

as he steals your wallet, he notices your nice Glock. But, before taking it home, he shoots you with it...just to make sure it works.

Most street assaults do not come with warning. Most happen from the rear. There's a good chance your holstered weapon will get used against you.

[TheWraith]

7. Actually, more accurate studies show that almost never happens.

Statistics like the one stated in the OP are inaccurate because they include cases of gang violence, where all parties are armed.

An FSU researcher named Gary Kleck (who is, incidentally, a card carrying ACLU member and Democrat) is widely considered to be the foremost expert in the US on defensive firearm use stats: he tells anyone who asks that the incidence of a gun being used against it's owner is almost nonexistant.

[Euromutt]

38. Ah, teachings from the College of It Stands To Reason

In your hypothetical scenario, the mugger just casually went from a first degree robbery charge to a first-degree murder charge. "It happens"? Well, if so, I'm sure you can cite, say, five instances of the events you describe (mugger incapacitates CCW permit holder, finds gun, shoots CCW permit holder with that gun) occurring within the past two years. That shouldn't be too much to ask.

[GreenStormCloud]

75. Most street assault DO come with warnings.

But the victims usually don't heed the warning or don't take proper precautions, so they become victims.

In your hypothetical, the victim appears to be alone, and is not watching around him. If he were practicing proper situational awareness, he would be aware of the approaching mugger and turn to face him before he gets that close.

He would avoid possible ambush areas from which a mugger can hide in waiting for the victim come within striking range. If he does have to pass a suspect area, he does so watching the danger area and with his hand on his gun.

Your scenario happens only if the victim is oblivious to his surroundings.

[yellowcanine]

3. Here we go again.......

While I applaud these researchers for doing this research, I wonder at the value of it.
People who want a gun are certainly not going to pay any attention to it. It is human nature to think that what may be true on average is not going to apply to me as an individual.

[atreides1]

4. Okay

One study in a specific area, only shows that possession of a gun during an assault in that area can be questionable.

As for me I'll wait until they conduct studies in all 50 states.

[cliffordu]

5. They don't differentiate between assaulter and assaultee, do they???

I'm not seeing it if they do.

I would bet if you are being assaulted and wave your gun around you stand a better chance of being shot.

If you simply shoot your assaulter without a lot of hoo haw, maybe not.

Conversely, maybe people waving guns around while assaulting others invite someone else to pop a cap in their asses.


Don't flame me, I like machetes.

[droidamus2]

6. Makes sense

If I am being assaulted by somebody that has a gun and I try to pull out my gun then it becomes a case of who aims and pulls the trigger first and considering the person assaulting me would already have their gun out it seems reasonable to surmise that in most cases they are going to get off the first shot. Doesn't mean they will hit me or I won't get off a shot but a good chance all trying to pull a gun is going to do me is get me shot. Now in a home, burglary type situation, where I might get the jump on the criminal I might have a slightly better chance.

[OKIsItJustMe]

13. According to the study, that scenario (armed resistance) increases the odds even more

http://www.ajph.org/cgi/reprint/AJPH.2008.143099v1.pdf

…

After we adjusted for confounding factors, individuals who were in possession of a gun were 4.46 (95% confidence interval =1.16, 17.04) times more likely to be shot in an assault than those not in possession. Individuals who were in possession of a gun were also 4.23 (95% CI =1.19, 15.13) 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 5.45 (95% CI =1.01, 29.92) times more likely to be shot.

…

[Fire_Medic_Dave]

43. Of course they included those in illegal possession of firearms.

[safeinOhio]

21. Does that include sexual assault?

Don't resist and you won't be killed, may be.

They may be correct, but in most studies you have to wait for the peer review to understand it.

[droidamus2]

31. Excuse me

I didn't make any suggestion on whether you should or shouldn't resist I was just pointing out that the study makes sense if you look the situation where both the victim and the perpetrator are armed.

[TPaine7]

34. What you say does make common sense...

but common sense sometimes varies from reality.

There are many factors that affect who actually has the upper hand. Is the person holding the gun on you perfectly attentive? The slightest loss of attention can change the equation, and the "victim" can change the attention equation. The victim can sometimes wait until there is naturally occurring attention loss.

Is the perpetrator really committed to killing if necessary? If not, the "victim's" commitment (backed by a clear conscience, and righteous indignation) can be devastating. The victim didn't enter the encounter determined not to do hard time for murder.

How hard is it for the "victim" to draw? If the draw is easy or well practiced, the "victim" can have a distinct advantage due to human reaction time. Think about it like this:

You have a baseball in your hand and you are threatening to bean me with it if I don't comply. I am very near to a baseball that is out of your line of sight. I engage you in conversation about your demands and then unexpectedly pick up the baseball and throw it full force at your head. Who will likely win this encounter (assume we are both equally skilled Major League pitchers)?

It's like a black belt Karate instructor explained to me years ago in college. If you are skilled in martial arts and a person is holding a gun or knife on you, you have the advantage--for the first move. It will take time for their brain to register the attack as long as you don't telegraph it. It's simple physiology.

The counterattack is your real danger. Thus, your first move must be devastating and incapacitating, either psychologically or physically--preferably both. It must remove them from the fight, or at the very least it must remove the weapon from the equation. (Something like broken, bleeding fingers on the gun hand and the gun on the ground near you might work in this situation.)

[Euromutt]

73. It's reasonable to surmise if you're unfamiliar with the mechanics of shooting

Hypothetical situation: an armed robber has the drop on you. He's got a handgun in his hand, with his arm extended towards you. What happens if you sidestep?

Naturally, he's going to want to keep the gun pointed at you, which means he has to swing his arm, which has a two-pound metal weight (to wit, the gun) on the end. He's got to overcome inertia to get the gun moving, and he's got to overcome inertia to bring it to a halt when it's trained on you again. And chances are, he's going to swing too far, so then he's going to have to swing the gun back, and again try to stop it when it's pointed at you.

In the interim, you have the opportunity to draw while sidestepping, bring your weapon to bear on him, and shoot. He may shoot in the interim, but his chances of actually firing while the weapon is aimed at you are slim while the thing's swinging all over the place. But because you're presenting your weapon directly towards where he's standing while he's swinging his gun around, you have no such problem.

Lesson courtesy of Rob Pincus.

[GreenStormCloud]

74. Add to your good post: Action beats reaction.

BG(Bad Guy) has the gun on GG(Good Guy)and has not fired. GG decides to fight. It is actually possible for a trained person to draw and fire before the BG can fire. In fact, it is fairly common. The BG has to realize that the GG is drawing, process that information to a decision to fire, then send the signal to his hand to pull the trigger. His brain has to do all of that just to catch up to where the GG's brain WAS 1/3 second ago. With a pocket gun, the BG's first information will likely be when he actually see the GG's gun, if he sees it.

If the GG has his gun in his trouser pocket he can claim to be getting his money clip while he is actually getting a gun. If he combines that with a seemingly nonthreating distractor move, then he can shoot before the BG knows what has happened.

Or the GG may shoot from within the pocket. Then the BG gets no warning at all.

Add your sidestip to that and the odds can shift dramatically to favoring the well prepared GG.

Add one other thing. Street criminals are almost always horrible shots.

[divideandconquer]

9. Why is the gun lobby afraid of research?

"which in the past have not been legally permitted to fund research ‘designed to affect the passage of specific Federal, State, or local legislation intended to restrict or control the purchase or use of firearms.’”

[X_Digger]

16. You mean "research" with the result pre-determined?

Shit, the quote lays it out baldly for anyone to see-

"research 'designed to affect the passage of specific Federal, State, or local legislation intended to restrict or control the purchase or use of firearms.'"

Read that again.. research designed to affect passage of restrictions. Not scientific inquiry, research _designed_ to a specific non-scientific aim- a political one.

wtf

[TPaine7]

35. +1

Or plus infinity.

There really isn't much left to say after that. This is a politically motivated, nonscientific study by a dishonest person operating outside of his field of expertise.

What more need be said?

[DissedByBush]

64. Afraid of biased research affecting policy

I haven't yet seen one of these that hasn't been easily torn to shreds.

The one that tried to answer the question "What if one of the students at VA Tech had a gun to shoot back?" was set up to be completely unrealistic and biased against the armed student.

[GreenStormCloud]

76. Actually, we welcome GENUINE research.

But this piece of Kellerman v2.0 just isn't it.

[Ian David]

10. Did they break it down by legal vs non-legal gun owners? n/t

[RamboLiberal]

11. I'm calling BS on this study

From what I'm seeing they didn't separate legal vs. illegal owners. I'm betting many of those shot were criminal element vs. criminal element. And it also looks like this study was done in a strictly urban area.

[DonP]

25. Isn't that the big issue with the Kellerman Study too?

IIRC Kellerman did his study in a densely populated urban area, with a high crime rate to begin with and at the peak of the crack cocaine tidal wave.

His study (the one that measured how much more "likely" you were to be shot by a gun in your own home) was skewed heavily by the sampling. I'm suspicious of the background on this - appearing just as SCOTUS accepts the Chicago case too.

Wow, what are the odds, huh?

[Euromutt]

50. It has exactly the same problems as Kellermann's 1993 study

The fundamental element of the protocol is the same shitty one as Kellermann's, namely comparing a study population of shooting victims to a control group of people who haven't been shot. That's your first confounding factor right there: your study subjects are by definition more likely to have been shot (chance of 1, in fact), because that was what you selected them for. Similarly, your control group are less likely to have been shot (chance of 0), because that was what you selected them for.

So you've got one bunch of people with bullet holes, another bunch without, and essentially you're asking "why do these guys have holes while those guys don't?" There's a whole raft of reasons why, but you're trying to isolate how much effect one of those factors has. For this, researchers use a technique called "multiple regression" in an effort to compensate for all the other variables, but the fact is that you can never be 100% certain that you actually have identified every variable.

That's why the validity of econometric modeling (which is what this study is) can only be tested in one way: its ability to make predictions better than chance. That's why a single econometric study is worthless; unless the model can be used to make predictions that stand up to testing, it is evidence of exactly nothing. In the words of Ted Goertzel (http://crab.rutgers.edu/~goertzel/mythsofmurder.htm):

When presented with an econometric model, consumers should insist on evidence that it can predict trends in data other than the data used to create it. Models that fail this test are junk science, no matter how complex the analysis.

Junk science is what Kellermann's work is, and it's dollars to donuts that it's what this study is too. Especially because the press release makes claims eerily similar to Kellermann's, and in Kellermann's case, even a cursory reading of the actual study showed that its findings really didn't support the authors' conclusions, and the press release seemed to have been written about a completely different study.

[Euromutt]

63. Oh, I forgot the big one: which way does the causal relationship run?

The problem with Kellermann's conclusions was that they weren't supported by his findings. The fact that he found a correlation between keeping a gun in the home and being the victim of a homicide didn't mean the former caused the latter. Were people more likey to be shot because they kept a gun in the house, or did they keep a gun in the house because they thought they were more likely to be shot? Given the higher rates of high-risk behavior among the study group (i.e. the dead guys), the latter explanation is more plausible.

We have the same question with this study: are people more likely to be shot because they carry a gun, or do some people carry a gun because they do things that are likely to get them shot at?

[X_Digger]

12. Open questions re methodology..

- Case control correlation (do the 677 & 684 participants match up for demographics like previous crime record, illegal drug use, income and neighborhood crime level)
- Were the victims shot with their own gun? (ie, did they factor out suicides since they're focusing on assaults?)
- Doesn't appear to be any discussion of non-shooting defensive gun uses

[OKIsItJustMe]

15. This addresses some of your concerns

http://www.ajph.org/cgi/reprint/AJPH.2008.143099v1.pdf

…

However, compared with control participants, shooting case participants were significantly more often Hispanic, more frequently working in high-risk occupations, 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 (Table 1).

…

The others I believe are addressed here:
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=118&topic_id=256837&mesg_id=256846

[X_Digger]

18. So the case control correlation wasn't legit.. gotcha.

They should have kept calling until they found another 35yo, white, male junkie to compare to.

[OneTenthofOnePercent]

22. Said otherwise: Shooting cases were predominantly gang related. Gotcha. Awesome study!

:sarcasm:

[Fire_Medic_Dave]

44. Many shooting case participants were likely engaged in illegal activity in illegal possession...

of a firearm. That couldn't skew the statistics at all. :eyes:

[Euromutt]

72. So the "shooting case participants" significantly fell into demographics unlikely to have land lines

...but the control group was established by randomly calling numbers in Philadelphia. Well, not quite randomly, right? Because federal law prohibits making unsolicited calls to cell phones, which means the control group by definition had land lines, whereas the shooting case participants were statistically likely to be members of demographics less likely to have land line phones.

That, right there, makes it highly unlikely that the control group bore any demographic resemblance to the shooting case group, which makes any comparison between the groups--and any resulting conclusions--meaningless. That fact alone invalidates the findings of this whole study, let alone the conclusions and press release.

Hereby hangs a tale: I worked as a caller for a market research company for a while back in the 1990s, and back then, it annoyed the piss out of me that the statisticians who drew up the questionnaires refused to take feedback from the callers regarding the quality of their work. In particular, the length of the questionnaire matters, because if it goes on for too long, those respondents who have something better to do are going to hang up before you're done. And when you only incorporate completed questionnaires into your data, it means you're in effect introducing a bias into your sample, skewing it towards those demographics who have the time to yak to a total stranger for half an hour or more. And this is something no degree in statistics will teach you; you learn this lesson only by conducting interviews "hands on," which is something very few researchers bother to do.

[Will E Orwontee]

23. Hardly first of its kind and actually contradicts USDoJ studies . . .

USDoJ says that persons resisting assault with a gun experience the smallest incidence of being injured, lower than persons using knives, baseball bats or golf clubs and even lower than for those who offer no resistance. Of course their stats come from actual crime reports of victims with known outcomes, not via econometric magic.

"On average in 1987-92 about 83,000 crime victims per year used a firearm to defend themselves or their property. Three-fourths of
the victims who used a firearm for defense did so during a violent crime; a fourth, during a theft, household burglary, or motor vehicle theft.

A fifth of the victims defending themselves with a firearm suffered an injury, compared to almost half of those who defended themselves with weapons other than a firearm or who had no weapon."

Guns and Crime: Handgun Victimization, Firearm Self-Defense, and Firearm Theft

NCJ-147003
April 1994
U.S. Department of Justice
Office of Justice Programs
Bureau of Justice Statistics

[Treo]

26. What Would denny Crane Do?

http://www.youtube.com/watch?v=j0q4EkQrZrY

http://www.youtube.com/watch?v=_KvO-8IvoCI

[TPaine7]

28. This study helps resolve the long-standing debate

about whether anti-gun "academics" are fools or dishonest. Charles C. Branas, PhD, Associate Professor of Epidemiology gives evidence of being both.

First, why would a epidemiologist do a study on the efficacy of guns in defensive use? Would you be interested in a study by Yo Yo Ma on the transmission vectors of H1N1? After all, Yo Yo Ma is a world-class expert.

“This study helps resolve the long-standing debate about whether guns are protective or perilous,” notes study author Charles C. Branas, PhD, Associate Professor of Epidemiology. “Will possessing a firearm always safeguard against harm or will it promote a false sense of security?”

That last sentence is classic sophistry. "Will possessing a firearm always safeguard against harm...?" Uhhh, duh. Of course not. Everyone knows that.

"Will possessing a firearm always safeguard against harm or will it promote a false sense of security?" Those are the only two choices. It's as legitimate as if someone asked you whether you killed your wife on Wednesday or on Thursday.

This man is dishonest and very biased. Furthermore, we have no reason to believe he is remotely qualified to conduct such a study. His "study" is not worthy of serious attention.

[OKIsItJustMe]

29. Sophistry you say…

You mean as opposed to constructing an analogy along the lines of, “Would you be interested in a study by Yo Yo Ma on the transmission vectors of H1N1?”

There's no reason to believe that a musician has any particular expertise in disease transmission.

There's also no reason to believe that an epidemiologist would have any particular expertise in the use of a cello, or a gun.

On the other hand, an epidemiologist does have an expertise at analyzing risks to human health.

[TPaine7]

32. Yes, sophistry...

First there was the sophistry I cited.

Second there was the sophistry of your calling attention to my use of the word sophistry, without the slightest attempt to refute--or even address--it. (A wise sophistry as sophistries go, as any such attempt would fail.) What I called sophistry was indeed sophistry.

Third, there is this last sophistry, this attempt to imply that expertise in analyzing risks to human health is the same as estimating the efficacy of using guns. The study of the efficacy of guns in defensive situations is not an analysis of risk to human health any more than the study of the efficacy of Firestone tires on wet pavement is an analysis of risk to human health. Yes, there is a connection--both affect human health. There is also a connection between music and the analysis of transmission vectors--both are mathematically grounded.

epidemiology

...

n.

The branch of medicine that deals with the study of the causes, distribution, and control of disease in populations.

http://www.answers.com/topic/epidemiology

1) Whether guns are useful in self-defense is not a medical question. Epidemiology is a branch of medicine.

2) Whether guns are useful in self-defense is not a disease related question. Epidemiology is a branch of medicine "that deals with the study of the causes, distribution, and control of disease in populations."

It would be as legitimate for this dishonest epidemiologist to study the efficacy of Firestone tires on wet pavement--another non-medical, non-disease based question that involves emergency room visits. His "expertise" would be just as authoritative.

[OKIsItJustMe]

40. It would be as legitimate for this…epidemiologist to study the efficacy of Firestone tires…

I agree.

I believe that statistics could be used to correlate the use of particular brands & models of tire on wet pavement with various accident outcomes; and that an epidemiologist would have the skills to do that analysis.

[TPaine7]

48. Let's put my comment in context

It would be as legitimate for this dishonest epidemiologist to study the efficacy of Firestone tires on wet pavement--another non-medical, non-disease based question that involves emergency room visits. His "expertise" would be just as authoritative.

In context, I am saying that his "expertise" in the efficacy of defensive gun use based in a PhD in epidemiology is just as valid as his "expertise" on Firestone tires. His degree is irrelevant. The question is not medical.

The relevant facts, as revealed in this thread are:

1) The author is dishonest--he used a false dichotomy to advance his agenda.
2) The study was intended to drive politics--not exactly a scientific motivation.
3) The study included gun use by criminals, drug users and convicts, and thus is not representative of gun use advocated by gun rights supporters.
4) The study was authored by someone working outside his area of expertise.

So while I agree that "statistics could be used to correlate the use of particular brands & models of tire on wet pavement with various accident outcomes" it is irrelevant in this case. A study conducted by an automotive tire expert would be bogus if it had similar flaws.

[OKIsItJustMe]

57. I have to ask

What make you think you can find the flaws in a study you have not read better than the peers who reviewed it?

What expertise do you bring to bear?

[TPaine7]

58. Hmm

What make you think you can find the flaws in a study you have not read better than the peers who reviewed it?

Nothing.

I didn't find flaws in the study, which you correctly state that I haven't read. I found flaws in the author.

1) He is dishonest.

2) He is attempting to leverage expertise to make himself seem authoritative in an area well outside his education. If he had presented himself as an amateur criminologist or at least not cited any medical expertise, I would not have made that criticism.

I typically will not spend a lot of time on a study by a dishonest author who misrepresents himself, so I didn't actually read the "study." I did, however, read the comments and accept the views expressed by some posters.

I did nothing that indicated that I was more qualified than the peers to find flaws in the study. However, now that you mention it, I am more qualified than they in this case. That is because in this case the bias is so egregious that basic honesty and a modicum of scientific knowledge are all that is required. I have both; they have medical knowledge and a passion for controlling the purchase or use of firearms.

As X_Digger pointed out in post 16, this "study" was "designed to affect the passage of specific Federal, State, or local legislation intended to restrict or control the purchase or use of firearms." They might as well have printed a disclaimer:

THIS POLITICAL PROPAGANDA PIECE IS NOT SCIENCE NOR SHOULD IT BE CONFUSED WITH SCIENCE IN SPITE OF ITS BEARING SOME OF THE TRAPPINGS OF SCIENCE AND APPEARING IN AN OSTENSIBLY SCIENTIFIC JOURNAL.

When there is BS like that on the surface, I rarely go as deep as Euromutt did in post 37, not because I can't but because I already know the answer. But his comments are quite revealing regarding the Kellermann debacle and the reliability of the "peer review" of medical propagandists.

What expertise do you bring to bear?

A nose for bullshit. A little knowledge of the world I live in. A modicum of intelligence. Time spent every day doing my own thinking.

That will suffice for a great many things in this world. For instance, if a guy asks me if I want to make a 1,000% return on my money while vacationing in Vegas, I say no thanks. If a lawyer asks if I committed the murder on Wednesday or Thursday, I know enough to say "objection, assumes facts not in evidence" even if my lawyer sits mute. And if someone gives me a false dichotomy, like the turd this author foisted on his readers, I know my intelligence has been insulted.

I think you may have misunderstood my "expertise" critique. Credentials are not necessary to profound contributions. A patent clerk can discover new laws of nature. If he presents his findings based on their scientific merit, all is well. If he attempts to bolster his credibility by citing his patent expertise, well, use your imagination.

[Euromutt]

62. A single study is evidence of very little

A single study using econometric modeling--like this one--is evidence of nothing at all. The key to determining the scientific validity of any study is whether its findings can be replicated by others, and the value of an econometric model depends on its ability to make predictions that are more accurate than chance. An econometric model that is incapable of predicting trends in data other than the data used to create the model is worthless.

And that's even if the science is impeccable.

[Euromutt]

65. Additionally, you don't need to be an expert on fashion design and tailoring...

...to see that the emperor's naked.

See "The Courtier's Reply": http://scienceblogs.com/pharyngula/2006/12/the_courtiers_reply.php

[GreenStormCloud]

33. I see one serious flaw in the study.

It only looks at gun on gun encounters. It does not look at encounters in which the criminal has a knife, or club, or just fists, or multiple assailants. Neither does it look at situations in which the criminal runs away - no shots fired.

I know several people who have defended themselves using a gun, and the criminal ran away, no shots fired. Even if they had lived in the area of the study, none of their incidents would have been included because nobody got shot.

The study is a biased, worthless, piece of political crap.

[SteveM]

51. Among other flaws in this study, these you mention stand out (nt)

[Tejas]

36. Does this include stripping firearms from LEO's?

It happens.

[GreenStormCloud]

39. This study tosses out 99%+ of successful DGUs.

Criminal is in beginning stages of attack, "victim" spots impending attacker, "victim" flashes gun, would be attacker breaks off attack, no shots fired. Since there is no shooting, it doesn't make the report.

This report is nothing more than gun-grabber propaganda.

[X_Digger]

42. Coming to a theater near you!

It's Kellerman II, the Sequel. See the continuing saga of our hero, the epidemiologist, as he tries to navigate the perils of criminology without a net! Thrill as he jumps to conclusions so far from fact that he can't see the other side! Chill as he leverages data into the form that he had already laid out.

We'll sell you the whole seat, but you'll only need the edge!

Playing now at the Hawthorne at 1:25 3:10 4:45 5:20 7:00 8:40 9:00 and a special 1/2 price showing at midnight.

[PavePusher]

45. Funniest. Thing. All. Week! n/t

:rofl:

[GreenStormCloud]

52. Love it. LOL. N/T

[Fire_Medic_Dave]

41. Did they look at legal owners or did they include those in illegal possession?

Let me guess. They included those in illegal possession. I'd be willing to bet $50 dollars on it.

[Euromutt]

47. In further news, "Insulin of questionable value in treating diabetes"

In a study conducted among test subjects who injected themselves with insulin, researchers found that all of them had developed Type I or Type II diabetes.

(Say it with me, folks: correlation does not imply causation.)

[tortoise1956]

49. I gotta call BS on this...

The study itself points out the main inconsistency - the the sample dataset was predominantly from a group that doesn't begin to represent a random sampling of the population of Philadelphia, while the control group was. That invalidates conclusions drawn between the two sets.

Another big issue is that this was based on a sample size of 6% of a group of 677 individuals, or 40-41 individuals. It's been a long time since I dealt with statistics of this type, but I would hesitate to draw any conclusions based on a sample size of 41, drawn from an extremely small subset of the general population. I certainly would smack any of my analysts who tried to pass off a comparison between this small subset and random individuals of a city as having any validity.

Ok, I'm done. Now it's time to go play with my metallic phallus...

[chibajoe]

53. Wait a minute, if I'm reading this correctly

they don't differentiate between victim and aggressor. What they are saying is that during a crime where both parties are armed, the chances of someone getting shot is 4.5 times greater. Well, duh.

[slackmaster]

59. Indeed, the sample set "Philadelphia residents who were shot in an assault" includes both

Victims and attackers.

[JonQ]

55. I'm guessing this is another one of those studies where they had the answer

then sought out evidence to back it up.

I'll have to read it more later, but these things are typically full of misdirection (being generous there).

[benEzra]

56. Hmmm, take a survey of criminals treated at an urban hospital

and then make axiomatic statements about lawful owners and licensed carry. One wonders how that passed peer review, but the bar seems much lower on gun-related topics.

Looking forward to the companion article in the Yale Law Journal, in which criminologists administer a survey to tuberculosis sufferers and then make opinionated statements about the epidemiology of H1N1 influenza...

[proteus_lives]

60. Horseshit.

That is all.

[Howzit]

61. Are we talking about people carrying guns illegally being shot by their peers?

How does this predict the outcome for the 99% of gun owners who follow the law?

[Taitertots]

66. Total Fail!

Anyone who defends themselves with a gun is not part of the study. They randomly selected 677 people who have already been shot. If you used a gun to defend yourself you are NOT TAKEN INTO ACCOUNT. They don't even attempt to differentiate between legal and illegally carried guns.

[JonQ]

67. Wow, I didn't catch that

what a terrible study.

You'd think the grabbers would be embarrassed that they have to resort to these tactics.

[spin]

68. Gun grabbers lie all the time...

That's why they are doing poorly in the national polls.

"You can fool some of the people all of the time, and all of the people some of the time, but you can not fool all of the people all of the time."
Abraham Lincoln (attributed)

[Indy Lurker]

69. Is this what passes for medical research?

It's sure not science!!!

"Penn researchers investigated the link between being shot in an assault and a person’s possession of a gun at the time of the shooting. As identified by police and medical examiners, they randomly selected 677 cases of Philadelphia residents who were shot in an assault from 2003 to 2006. Six percent of these cases were in possession of a gun (such as in a holster, pocket, waistband, or vehicle) when they were shot.

These shooting cases were matched to Philadelphia residents who acted as the study’s controls. To identify the controls, trained phone canvassers called random Philadelphians soon after a reported shooting and asked about their possession of a gun at the time of the shooting. These random Philadelphians had not been shot and had nothing to do with the shooting. This is the same approach that epidemiologists have historically used to establish links between such things as smoking and lung cancer or drinking and car crashes."

Based on the 4.5 times more likely, can I conclude that only 1.33% of the general population carry a firearm?

1.33% of the general population carry, 4.5 times as many (on a percent basis) or 6% of the assault victims who were shot carried firearms.

Therefore carrying a firearm is linked to being shot during an assault?

They think this is valid because x% of the general pop smokes, and y% of lung cancer victims smoke (and Y% is much greater than x%) smoking is linked to lung cancer?


One is a disease, one is a behavior. There is no reason to assume such a study would be valid. This was a waste of money.

One can easily think of other behaviors with a common "link" that is completely without basis.

It's well known that students in urban areas have higher dropout rates, but it would be foolish to attempt to "link" the high dropout rates to simply living in an urban environment

Let's see how silly this sounds "The study estimated that (students) with (an urban zip code) were 4.5 times more likely to be (dropout of school) than those not possessing (an urban zip code).

If only they would move to a different zip code, they would finish school.

************************************

Some other thoughts

They only look at victims who were shot. Doesn't this severely limit and skew the results. What about all the assaults that were defeated by a firearm without anyone being shot, Where do these figure in?

Of people shot during an assault 6% had firearms, and 94% did not. Do they conclude it's safer to NOT have a firearm?

Does this indicate police would be safer if they disarm?

[PavePusher]

70. If this is considered "medical research"...

It sure explains a lot about the pharm industry...

[Euromutt]

71. Not medical research, "public health research"

That's where medical personnel (who are not necessarily scientists by training) dabble in the world of econometric modeling, and make an even greater hash of it than social scientists.

The comparison of this study to research linking smoking and lung cancer is--as you rightly note--spurious (even leaving aside the fact that a term like "associated with" means nothing more than "a statistical correlation exists though a causal relationship has not been established," while the term "linked to" is practically meaningless). For starters, once researchers started looking into lung cancer, they found that something in the order of 85-90% of lung cancer patients were smokers, and smokers had statistically significant chances of developing lung cancer, and these chances increased the more and longer a person smoked (e.g. a person who smokes 40/day for 40 years has ~25% chance of developing lung cancer).

Now, when you've got that kind of breakdown, that's hard to argue with. But there's nothing comparable that emerges from this study. For starters, of the shooting victims, you've only got 6% carrying a firearm, against 94% who weren't. Obviously, that's not the whole story, since in 2006, the population of Philadelphia county (which is coterminous with the city) was ~1,448,000 (US Census Bureau estimate) while the number of issued Licenses To Carry was 5,957. But since the researchers didn't bother to make the distinction between legal carriers and illegal carriers, this number doesn't help you analyze their results.

But then you have to wonder why they didn't make that distinction, because their conclusions and press release most assuredly make recommendations aimed at legal carriers. Hell, I wouldn't be surprised if this is Kellermann's bullshit 1993 study all over again, where the "homicide victims" turned out (after he finally published some of his research data after several years) to include armed criminals shot by local law enforcement. And Kellermann's study only covered people killed in their own home; how many of the "victims" in this study were, say, armed robbers or drug dealers shot (not necessarily killed) by Philly PD?

That's pretty much your litmus test right there: if you examine that part of the protocol that can be gleaned from the abstract and the press release, does it bear any relationship to the researchers' conclusions?

There's another factor occurs to me, based on the problems identified with opinion polls during last year's presidential election, which is this:

To identify the controls, trained phone canvassers called random Philadelphians soon after a reported shooting and asked about their possession of a gun at the time of the shooting.

Given federal restrictions on making unsolicited calls to cell phones, this means the control group consists entirely of people with land lines. Now, as OKIsItJustMe has helpfully quoted:

However, compared with control participants, shooting case participants were significantly more often Hispanic, more frequently working in high-risk occupations, 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 (Table 1).

That's almost a perfect summary of demographics unlikely to have land lines. So you've got a control group that is almost certainly highly divergent from your study group. Bravo, you've just invalidated your study.

[benEzra]

78. Oh, whoa, wait a minute. The control group consisted of calling random people

and asking them if they were in possession of a gun at the moment? Ummm, did they correct for false "no" answers, and if so, by how much? Not to mention the built-in biases you mention.

Personally, if my landline rings and a stranger on the other end says claims to be a researcher and asks if I have a gun on my person, the answer is "no." Period. Even if the answer is yes.

Thing is, if a "study" of HIV infection rates called random people out of the phone book, asked whoever answers if they are HIV positive, and called that a "control group," they'd be laughed right out of peer review.

[Euromutt]

79. That was a problem with Kellermann's 1993 study too

All it would have taken was for a comparatively small number of the controls to have falsely denied owning a gun to render the association statistically insignificant. And the controls were better matched to the case subjects than in this new study (i.e. they weren't polled by randomly being phoned).

[spin]

77. The researchers didn't want to consider...

assaults defeated by a firearm.

If they had, the conclusion would have have different.