A few year end roundups for all of y’all.  My big plans involve champagne and my pajamas, and a whole lot of gratitude for a wonderful year.

First, 5 retracted research studies.  #2 is weirdly hilarious. (h/t Maggie’s Farm)

Then, best data shares of 2012.  Way to represent Boston!

This list is a bit more modest….not “best of” rather “20 great” infographics.

And here are about 10 different “best science books” links, all in one place.

Unrelated to data, but good for a laugh:  Dave Berry’s year in review.

That’s all for 2012.

Don’t forget that 2013 is The Year of Statistics!

Happy New Year!

This always makes me smile.

In my last post, I took a look at the idea that gun fatalities were set to surpass traffic fatalities as a cause of death in the US.  In the comments, SJ pointed out that the word “surpass” was not appropriately use.  The primary driver of this convergence was really the decrease in traffic fatalities, so it would have been better phrased as “traffic fatalities set to drop below firearm deaths”.  The article I originally cited had this to say about the drop:

The national gun-death rate would not be approaching that of motor vehicles if it weren’t for the fact that the latter has dropped fairly drastically in the past half decade or so thanks to an increased effort to make the nation’s roads and vehicles safer. Gun-rights advocates will point to the relatively subtle rise of the gun-death’s purple line to argue that we don’t need to pass more gun restrictions. Gun-control advocates will point to the more severe drop of the yellow line to make the case for what might happen if we were to.

It got me curious though….was that drop really do to some new law or set of regulations?  I’ve been a licensed driver for over a decade now, and I couldn’t recall any specific big changes in the past few years that change how I drive. Let’s take a look at that chart again:

• Be cautious in assuming that a sudden, large drop in fatalities is in response to interventions related to vehicle design. It takes about 20 years to turn over the fleet. 

• Don’t expect most regulatory actions aimed at drivers to produce a sudden, huge drop in fatalities because such actions usually target only a portion of drivers (such as improvements in graduated driver licensing targeting young drivers only). 

• Realize that any sudden, large reduction in fatalities is likely only an unintended byproduct of factors that influence the entire transportation system, such as a rapid change in the economy. 

• Be aware that most rapid, underlying changes are transient, and therefore, their effects are mostly transient, too.

On Slate this morning there was a headline about how firearm deaths have surpassed motor vehicle deaths in 10 states.  The article cited a report by the Violence Policy Center that showed that between 1999 and 2010 there was a  decrease in motor vehicle deaths nationwide, and a slight increase in firearm related deaths.  Thus, the report and article argued, we are heading towards a day when firearm deaths surpass motor vehicle deaths.

I’ve blogged before about how deceptive I think the term “firearm deaths” is (as it lumps in police action, homicides and suicides as one group), and it turns out this is the same report that annoyed me the first time.  This time however, I was struck by the chart they included in the article:

Essentially, homicides by gun have not changed in the 11 year period covered here, though they did tick up in the middle.  Suicides have gone up.  If you’re curious what the other three categories look like when they’re not all bunched up:

So unintentional fatalities are decreasing, and the other two categories appear to be fairly constant.

As a reference point, the population was 280 million in 1999, and was 309 million in 2010.

To be honest, I don’t have a strong opinion about gun control, but I do hold data integrity at a premium.  If we’re going to talk about regulating guns in order to keep people safe, you have to either include suicide prevention as one of your foremost points, or you have to start using the homicide data only.  I do not believe it is intellectually honest to quote total firearm death statistics when the national conversation is clearly focusing on homicides.

If we don’t start with the right data, how will we know if any interventions actually work?

I got in to an interesting discussion today over at sciencebasedmedicine.org about medical errors and the huge variety of numbers that are out there.

It reminded me of a project I started to work on a few years ago at work, where I attempted to correlate low staffing levels to an increased error rate.

When I finished the graph, I discovered that the trend was completely in the wrong direction….low staffing levels actually appeared to be correlated with fewer errors.  When I did some fishing around, it became obvious what had happened.  The error reporting system is all self reporting by staff.  When they don’t have enough staff to cover all the work, they also don’t have time to fill out error reports, and thus many small to mid sized errors went undocumented.*  When there was more staff available, people were more compliant in their error reporting.

This is yet another risk of self reported data.

I will admit, it still amuses me that somewhere deep in my documents file, I still have a spreadsheet that suggests that the fastest way to get rid of errors would be eliminate most of the staff.  Correlation does not equal causation, and graphs are only as good as the source data.

*Hospital policy dictates that we should record “near miss” errors…..situations where nothing actually went wrong, but almost did.  The theory is that we need a record of these issues so we can address things before they actually cause a real problem.  However, if no one documents them at the time, they are nearly impossible to find later.

Hope everyone had a merry Christmas….mine was most definitely lovely.  I got an iPad and a garbage can full of manure.  I’m relatively sure I’m one of only a handful of people on the planet to get that combination of gifts (and love them both!).

Imagine Hamish McDonald, a Scotsman, sitting down with his Glasgow Morning Herald and seeing an article about how the “Brighton Sex Maniac Strikes Again”. Hamish is shocked and declares that “No Scotsman would do such a thing”. The next day he sits down to read his Glasgow Morning Herald again; and, this time, finds an article about an Aberdeen man whose brutal actions make the Brighton sex maniac seem almost gentlemanly. This fact shows that Hamish was wrong in his opinion but is he going to admit this? Not likely. This time he says, “No true Scotsman would do such a thing”.

I thought this was an interesting accusation.  If taken seriously though, I wondered how this allows anybody to protest any characterization of any group they happen to belong to?  I mean for the moment we’re talking about feminists, but nearly every religious/ethnic/cultural group has to defend themselves from people who claim that their group is inherently violent/bigoted/crazy.

What always gets me in these claims is that people want to keep with sweeping generalities rather than even admit the data they’d need to ever make that particular claim provable.  Are feminists out to destroy men?  Well, in order to ever assess that, you’d need to figure out approximately how many women define themselves as feminists (using a population representative sample of course) and then figure out how to assess whether or not they wanted to destroy men.  This would get you in to assessing “hidden” agendas as well, as many would likely not admit it, meaning you’d have to ask tricky questions to sort through it, and likely get some false positives.  That’s a lot of work.

Anyway, I’m not a philosopher, but I’d guess the “no true Scotsman” fallacy is one that should be hurled with care.  Group identity is any area where we will almost never have data to back up our feelings, and that needs to be kept in mind.

If that’s what you celebrate of course.

No big post today, as I’m cooking Christmas brunch, but I did want to highlight a few good Christmas gifts.

We didn’t go crazy on Christmas gifts for the little lord this year, opting instead for some books.   He’s just starting to get interested in books put in front of him (mostly to shove in his mouth) but I have a feeling by this time next year he’ll enjoy Dr Suess almost as much as I do (for my money “barbaloot suites” is one of the most fun phrases to say, possibly ever).

Also promising is the book Big Questions from Little People (and Simple Answers from Great Minds).  It’s the kind of book I loved as a kid, and in fact I’m reading it now (he doesn’t seem to mind that I borrowed it).

My favorite Christmas present I can remember was the Big Book of Amazing Facts, which I read so often the cover fell off.  Not actually sure that Amazon link goes to the right book….it seems to be a fairly common title, but the publishing date on that one seems about right.

If you want to see some more fun reminiscing about curiosity inducting gifts, see this NYT article today on “gifts that keep on giving, if not exploding“.

And most of all, happy holidays to you and yours.  Thanks for stopping by.

I know we’re all supposed to stop talking about the Mayans, seeing as how the world was supposed to end on Friday and today’s Monday and all….but this link was too good to pass up.

A fiddler on the roof. Sounds crazy, no? But here, in our little village of Anatevka, you might say every one of us is a fiddler on the roof trying to scratch out a pleasant, simple tune without breaking his neck. It isn’t easy. You may ask ‘Why do we stay up there if it’s so dangerous?’ Well, we stay because Anatevka is our home. And how do we keep our balance? That I can tell you in one word: tradition!

I didn’t used to care much about tradition.  Then I got married.

I don’t know exactly what happened, but somewhere on my way down the aisle, I got totally overwhelmed by the idea that I was about to partake in a ceremony that literally billions of people had gone through for thousands of years.  According to Stephanie Coontz,  there is only one documented culture throughout all of history that did not build it’s society on marriage (can’t remember the name now….though I believe it was somewhere in Asia and formed family structure based on siblings all living in the same household).  It was the most connected I’ve ever felt to the rest of humanity (until I had my son), both to those here and the faithful departed.

Since that time, I’ve always been a little dubious of those who want to evaluate cultural tradition as though those deeper roots shouldn’t count.  In case you can’t see where I’m going with this, this happens a lot in sexism research.  Recently, a study performed at UC Santa Cruz suggested that men and women still think men should do the proposing.  The most cited reason by the participants was “a desire to adhere to gender role tradition”.

This whole thing caused some hand wringing, with my favorite Slate writer, Amanda Marcotte commenting that it is “benevolent sexism that leeches women of much of their autonomy beyond just the right to say yes or no”.

Sigh.

Alright, here’s the thing.  We’ve kept the tradition, but the whole back story?  It’s changed.  Nobody gets engaged at random anymore(okay, someone must).  People talk about this stuff first.  Women even initiate a lot of these conversations.  The tradition has stayed, but a woman’s agency to ask where the relationship is going, to discuss rationally if marriage makes sense as a next step, has developed.  This study did not ask men or women what they envisioned leading up to the proposal, just what that one moment would look like. All the articles I read about this talked about “how little had changed” in the realm of marriage when it came to equality for women.  Poppycock.

The traditions haven’t changed, but all the behavior around them has.

One of my favorite advice columnists is Dan Savage, a liberal gay male who is most notable for creating Rick Santorum’s “google problem”.

I listen to his podcast, and he routinely gets callers that say things like “I’ve been living with my boyfriend for 5 years, and now I want to get married but I can’t propose because I’m really traditional so I don’t want to bring it up” or “my boyfriend of 2 months and I are having sex problems, and I think it’s because he’s a really traditional Christian guy”.  He gets straight to the point with these people:  a traditional Christian does not have sex with someone they’ve been dating for two months.  A traditional couple does not live together for 5 years before talking about marriage.  Those things were unacceptable for most of history, and you need to come to grips with that.  People love to call themselves “traditional”, but most of them don’t like to act like it.

So give the college kids a break.  Women have an unprecedented level of say over who they marry and when, they’re just trying to keep their balance.

I was going to post this yesterday, but then I thought maybe the world was going to end.

Oops.

Seemed like such a lock too.

So here we are, a day late, with some entertaining links for your viewing pleasure:

First a bit of holiday cheer from reader Andy….how many elves would Santa need on his staff to get it all done?  12 million, apparently.

In data news:  This week the Sunlight Foundation released a new data app called Sitegeist.  It downloads 5 pages worth of data about your surroundings, including demographic and housing data, political contributions, and interesting things to do.  I’ve been using it for a few days, and it’s pretty interesting.  Apparently my neighborhood donates twice as much to Republicans as Democrats.  Who knew.

Speaking of Republicans….are you conservative?  Do liberals tell you that means your anti science?  Here’s a fast retort….ask them what was up with this.

I’ve had a weird uptick in the number of designers emailing me asking me to share their infographics on my blog.  Don’t they know infographics are dangerous AND I HATE THEM?

All right, I’m a bit of a hypocrite, I did think the top 10 sports infographics of 2012 were pretty cool