Home > data science, modeling, news, open source tools > The Sandy Hook Project

The Sandy Hook Project

February 1, 2013

I wanted to share with you guys a project I’ve been involved with started by John Spens of Thoughtworks regarding data collection and open analysis around guns and gun-related violence. John lives in Connecticut and has friends who were directly affected by the massacre in Newtown. Here is John’s description of the project:

I initiated the Sandy Hook Project in response to this need for information.  The purpose of this project is to produce rigorous and transparent analyses of data pertaining to gun-related violence.  My goal is to move beyond the rhetoric and emotion and produce (hopefully) objective insight  into the relationship between guns and violence in the US.  I realize that objectivity will be challenging, which is why I want to share the methods and the data openly so others can validate or refute my findings as well as contribute their own.

I’ve put the project on GitHub. (https://github.com/john-m-spens/SandyHookProject).  While it’s not designed as a data repository, I think the ubiquitous nature of GitHub and the control enforced through the code repository model will support effective collaboration.

John has written a handful of posts about statistics and guns, including A Brief Analysis of Firearm-related Homicide Rates and Investigating Statistics Regarding Right to Carry Laws.

In addition to looking into the statistics that exist, John wants to address the conversation itself. As he said in his most recent post:

What limited data and analysis that exists is often misrepresented and abused, and is given much less attention than anecdotal evidence.  It is relatively simple to produce a handful of cases that support either side in this debate.  What we really need is to understand the true impact of guns on our society.  Push back by the NRA that any such research would be “political opinion masquerading as medical science.” is unacceptable.  We can only make intelligent decisions when we have the fullest picture possible.

John is looking for nerd collaborators who can help him with data collection and open analysis. He’s also hoping to have a weekend datafest to work on this project in March, so stay tuned if you want to be part of that!

  1. JSE
    February 1, 2013 at 8:39 am

    This sounds like a great project. My only criticism is that calling it “The Sandy Hook Project” is not contributing to his goal of “moving beyond the rhetoric and emotion.”

    • February 1, 2013 at 9:29 am

      Yes and no – as you may remember, NRA’s LaPierre took a very different lesson from that massacre. It means different things to different people.

      And by the way, you gotta read this, hot off the press:

      The NRA vs. America (Rolling Stone)

  2. February 1, 2013 at 9:24 am

    I’m interested!

  3. suevanhattum
    February 1, 2013 at 10:24 am

    Would he be interested in hearing from someone who disagrees and uses lots of data to support his stance? I’m thinking that might help him strengthen his procedures. Matt Springer at Built on Facts (a great physics blog) has (sadly, imo) written two posts arguing against gun control. His information is definitely worth looking at if you’re pulling together data on this topic.

  4. Micah
    February 1, 2013 at 2:12 pm

    Addressing the conversation is the most important part. In order for this project to help anybody move beyond emotion and (I’m guessing here) to the conclusion that guns-for-all is unhealthy, I suggest a basic hypothetical question with a more difficult answer:

    What sort of evidence would someone have to present to you in order to convince you that freedom of speech and freedom the press are unhealthy and we should drastically scale back the First Amendment?

    It’s one thing for someone in New York to look at some numbers and rationally tell someone in Greeley, CO that guns are not actually making anybody safer. But what if someone from another planet rationally presented evidence that freedom of the press is a drag on progress to the point of causing unnecessary suffering ? When they tell you that, no, a blogger with a laptop isn’t actually going to be the last straw between freedom and tyranny, would you stoically review the evidence and hand over your laptop?

    Of course, I’m all for freedom of the speech, to the point that I can’t even imagine what sort of evidence someone to have to give me in order to accept giving it up. Even in the face of the a preponderance of evidence, I would probably just say, “who cares? now please don’t destroy my culture.”

    • JSE
      February 1, 2013 at 3:11 pm

      This is indeed a good thought experiment to do. But it’s complicated in this case by the dissensus about what the 2nd amendment actually says. I’ll bet there are people whose stance is “the overall effect of guns is to increase violence and death, and I would never have a gun in my house, but there’s an inherent value in relaxing Constitutional guarantees, and among those guarantees is having as many handguns as you want in your house with limited government oversight, and so I oppose gun control and would oppose it no matter how sure I was that gun control would save lots of people’s lives.”

      But that isn’t the stance of most proponents of gun control — I think most would say that freedom of gun-owning is not in the same Constitutional category as freedom of speech or freedom of the press.

      By the way, you don’t have to go to another planet to find non-Americans who will tell you that they get along just fine without absolute guarantees of freedom of speech, and that they are completely at ease with certain kinds of speech that we consider sacrosanct being banned for utilitarian reasons. Ask a German about the freedom to be a neo-Nazi, or a Scientologist.

      • February 1, 2013 at 5:34 pm

        Micah’s post is neither good nor a thought experiment. It is a false comparison. Micah’s argument is equivalent to this one, which might be offered by an addict: “Even if there is evidence that heroin is harmful, it doesn’t matter because heroin is my culture. What if there were evidence that breathing is harmful? What then? Would you stop breathing?” (You might even make a ridiculous argument that breathing IS harmful, because it increases carbon dioxide levels.)

        My point is that Micah is advocating for disregarding evidence completely in a discussion of guns, on the basis of an unsupported claim that guns are somehow good or a right independence of evidence. This is ridiculous.

        • Micah
          February 1, 2013 at 8:00 pm

          C’mon.

          I’m not arguing for disregarding evidence, I’m asking a question about how strong and what sort of evidence is persuasive enough to have someone give up a huge part of their culture. You appear to arguing that we should give heroin addicts articles from NEJM.

          I’ll take your (hypothetical ) breathing example. Hypothetical meaning I don’t actually believe that there is a good alternative to breathing. But suppose there was, someone came along and offered me an insert that would come out of my head, and plenty of evidence that this would add 5-10 years to my life. I, for one, would have trouble making the switch. If you lived in a city were everyone had inserts coming out of their heads, it might be a no-brainer for you. But for me, I’ll go ahead and take my measly 80-85 years on this earth on not worry about the evidence against breathing.

          The best that a just-the-facts study could do is say something to the effect of the following: “with guns, your chance of being the victim of a violent crime is .18 percent, without guns your chance drops to .14 percent.” Now if you are a robot, or if you grew up in Manhattan, this is plenty of evidence. But if you grew up hunting and have lived your whole life thinking that individual gun ownership is at the core of America’s strength, you would be able to accept and dismiss the .04 percent difference without even bothering to argue against that data or the data analysis itself – you would argue against something much larger and intangible.

  5. suevanhattum
    February 1, 2013 at 5:17 pm

    If what this link says is true (and it seems to be), I’m shocked that I didn’t know this long ago. It seems the 2nd amendment shouldn’t hold the same place in our hearts that the rest of the bill of rights does: http://truth-out.org/news/item/13890-the-second-amendment-was-ratified-to-preserve-slavery

    • NotRelevant
      February 2, 2013 at 1:02 pm

      Thanks, the article was very interesting.

      We often forget how much the Constitution and Bill of Rights were consumed with issues related to slavery. In fact, my public education history courses taught that the issue of slavery nearly split the USA before it even started. Imagine where we’d be today if two separate countries had emerged out of the constitutional convention in 1789. Not such a good place, as North America would have been vulnerable to re-invasion from Britain or even France. And Abe could never have said things like “a house divided, bla, bla, bla.” Likewise, John Brown’s Raid in 1860 would have been an act of international terrorism, not a criminal offense.

      One thing the article overlooks is that the word “people” means individuals in all places where it is used in the Bill of Rights. To ignore its significance in “the right of the people to bear arms…” is to advocate.

  6. Micah
    February 1, 2013 at 8:31 pm

    Here’s a not-hypothetical question. Quantify the value of gun rights to this guy -> http://theprowersjournal.com/wp-content/uploads/2012/04/LCC-Rodeo-Boys-290×290.jpg.

    Not the value to you or me. But to him.

  7. suevanhattum
    February 1, 2013 at 9:47 pm

    But Micah, I don’t think anyone is talking about taking away the right to use a gun to hunt in the woods. Long guns that make one shot (or a few?) at a time, those aren’t the problem, right? Semi-automatics – things that can suddenly kill a room full of people – I think those are what we’re talking about. I think the hunters can help us figure out the boundaries here.

  8. franklinvera
    February 2, 2013 at 1:01 am

    I think this is a very nice project. It is good to have objective information. On the other hand, from my own point of view on the gun problem, I think it probably won’t be very useful. Let me explain myself. Suppose you have plenty of data that links guns to gun-related violence. It seems to be clear that you are also going to have a lot more data when looking at other means by which people die or get injured (cars, smoking, drugs, …). That is why the discussion on guns using this statistical rhetoric, I think, can’t produce winners.

    My own opinion is that the issue about guns is an ideological one. It has deep roots in that constitutional amendment which gives gun possession in the US a strong ideological support. Also historical roots, the conquest of the west, the wars in which US has been involved (in action or in trade of weaponry), the film industry, news. Guns, as an idea, are well inside, I think, American culture.

    Ideological legitimization of guns allows, for example, to having people that love their guns. As an idea, it can be promoted, respected, shared, frowned upon, disliked, etc. (I personally dislike it and don’t promote it but this is clearly not the case for everyone). I think it is more useful to have statistics about this, love of guns, what is linked to love of gun, what promotes it, etc.

    Moreover, and this could be confirmation bias, in Cuba, where gun are exclusively used by the army, police, and very very few individuals with gun permits, like hunters, judges or things like that, it is more common to find fists, knifes and machetes, as a mean of conflict resolution. What I mean is as an idea, culturally, beyond the actual statistics of how many people make attacks using fists, knifes or guns. In Cuba, as a cultural thing it is more common to see a person making a belligerent stance by rising the fist to their faces, like boxing. On the other hand, I have noticed here (and this here is Canada, I haven’t stayed in US long enough) it is common to find people that express belligerence by gesturing like they are reaching for a gun in their backs. Also, and this one really surprises me, I see people that walk or gesture or simply constantly have one hand inside their jackets, like holding a gun. I have seen this a lot! I

    This cultural association of gun as a conflict resolution, or even more irrational for me, the association of guns and love, is what I think is the most dangerous about the gun issue in US. I also think this should be the point that is critical to address in order to improve in the gun problem in US. This is just my opinion. Combined, of course, with some gradual legislation about gun control or gun freedom.

    This said. Having numbers ready is good. In one of those endless (and in my opinion pointless) discussions of pro-gun vs anti-guns in which statistic of gun violence are cited, one wouldn’t want to be one of the contenders and not bring some numbers to defend yourself. 

  9. FogOfWar
    February 5, 2013 at 7:25 am

    Thoughts about this: http://en.wikipedia.org/wiki/More_Guns,_Less_Crime (and the reaction to it)?

    There isn’t an absence of statistical analysis on this question, there’s an abundance of heated rhetoric on both sides.

    FoW

  10. February 12, 2013 at 1:20 pm

    What are you trying to do?
    Improve the science, or improve the policy and folks’ behaviors?
    My free advice: don’t mix the two – the first is about data and analysis; the latter is politics and marketing.
    I am not a strict utilitarian when it comes to science (I do believe it is important have science about a topic even if – especially if- the politics/marketing side seems immutable) but if this project is the former; I’d like to see the data collected be tested out against one or two example hypothesis that are new; including international comparisons.

    Some good examples can be found here: http://www.dailykos.com/story/2012/03/26/1077930/-Statistics-Guns-and-Wishful-Thinking

    Have you seen this US-CDC report? Comparing to other 25 industrialized countries, 86% of all firearm-related fatalities in children under the age of 15 occurred in the US.

    http://www.cdc.gov/mmwr/preview/mmwrhtml/00046149.htm

    I would posit the data is quite clear to date- household gun ownership in most settings in the US is a dangerous addiction (especially dangerous to women and children) that tends to decrease household safety. Now you need to move on to marketing and policy – and good data is just a minor accessory in doing a great job with those two.

    I would do data collection around finding positive deviants who have stopped owning guns they didn’t really need to see if the reasons and emotions behind that decision can be communicated and replicated in others. i.e. data for a great marketing campaign.

  11. February 12, 2013 at 1:22 pm

    What are you trying to do?
    Improve the science, or improve the policy and folks’ behaviors?
    My free advice: don’t mix the two – the first is about data and analysis; the latter is politics and marketing.
    I am not a strict utilitarian when it comes to science (I do believe it is important have science about a topic even if – especially if- the politics/marketing side seems immutable) but if this project is the former; I’d like to see the data collected be tested out against one or two example hypothesis that are new; including international comparisons.

    Some good examples can be found here: http://www.dailykos.com/story/2012/03/26/1077930/-Statistics-Guns-and-Wishful-Thinking

    Have you seen this US-CDC report? Comparing to other 25 industrialized countries, 86% of all firearm-related fatalities in children under the age of 15 occurred in the US.

    http://www.cdc.gov/mmwr/preview/mmwrhtml/00046149.htm

    I would posit the data is quite clear to date- household gun ownership in most settings in the US is a dangerous addiction (especially dangerous to women and children) that tends to decrease household safety. Now you need to move on to marketing and policy – and good data is just a minor accessory in doing a great job with those two.

    I would do data collection around finding positive deviants who have stopped owning guns they didn’t really need to see if the reasons and emotions behind that decision can be communicated and replicated in others. i.e. data for a great marketing campaign.

    • March 19, 2013 at 9:20 pm

      I agree about the importance of separating the analysis of what experimental policies for reducing gun deaths have produced apparent success, from what arguments would be most persuasive to get those policies implemented in the U.S. If you haven’t made up your mind what solution you think is appropriate for the problem before looking at the data, good for you – but you may be in the minority. So don’t assume the hard work that goes into identifying a promising solution will be very persuasive once you feel ready to try and sell that solution to policy makers, or voters.

      But I have looked long and hard at some of the scholarship on the vaccine-autism debate, and one of the most striking features of the literature on that topic is the shoddiness of most published comparisons between countries, in terms of scientific rigor. Evidently people who understand the strengths and weaknesses of the available methods tend to avoid “ecological analysis” altogether, and people who go in for an ecological analysis can therefore get away with saying virtually anything about the implications of a comparison between two or more countries.

      For instance, many peer-reviewed journals indexed in PubMed have carried articles on the autism debate guilty of over-interpreting data from comparisons between countries with a sample size of 2 (this country vs. that country), a research design choice that should drive most statisticians batty. It’s too likely that a correlated difference between only two subjects (here, countries) will be a chance finding, and that problem is often overlooked when the data on the two subjects come from large populations of constituents (the individual fatalities, in this case). Wherever you stand on the question of whether vaccines have any potential to effect the risk of a child developing autism, the sloppiness of some of the published research on both sides of the debate has been jaw-dropping.

      And even when the research is done more carefully, international comparison data are easy to misrepresent. For instance, the CDC report you cite has the benefit of doing some standardized statistical comparisons (i.e., comparing rates of gun deaths in children after correcting for differences in population size among the countries being compared), but a general audience would not see the problem with interpreting the 86% statistic you cite as a “smoking gun”. (The more obscure “rates per 100,000″ are a better metric for comparison, but when the summary statistic is something like “0.32 compared with 0.03″ it just doesn’t sound nearly as impressive.

      This is the best article I could find on short notice about the “ecological fallacy” and best practices in interpreting international comparison data: The ecological fallacy strikes back, by Neil Pearce.

      http://jech.bmj.com/content/54/5/326.full

  1. No trackbacks yet.
Comments are closed.
Follow

Get every new post delivered to your Inbox.

Join 1,814 other followers

%d bloggers like this: