Author: bs king

Friday Fun Links 1-28-13

/ bs king / 3 Comments

I’m trying to learn how to use Twitter now, so I’m still learning this whole hashtag thing.  However, I know enough to know I love #middleearthpublichealth

This guy changes music based in a minor scale in to music based on a major scale.  While it likely makes hipsters cry, I found it fascinating.
Statistics, it’s all fun and games until someone gets accused of using data to aid an international conspiracy against Greece.

Have I ever told you I’m one of the few people under 35 who actually knows how to use a slide rule?  Well I am.  Oh hey, here’s one now!

I don’t know why I found this so interesting, but here are the most searched for out of print books in 2012.  Good job Madonna, still #1.

Sweden and rape

/ bs king

I’ve written before about the dangers of comparing international data, but a recent stat floating around reminded me that it’s not just public health data that’s tough to compare.

There’s a stat going around about the rapid increase in Swedish sexual assault.  The article shows this graph:

Which purports to show a 500% increase in sexual assault…with a particular rise between 2004 and 2005 (I don’t speak Swedish, so if this graph is actually about something else, my apologies).

Anyway, I vaguely remembered around the Julian Assange case that there was some assertion that the Swedes had a particularly broad definition of rape, so I went digging to see if I could find anything about what was going on.

I found this report from the BBC, which included this tidbit:

“But the major explanation is partly that people go to the police more often, but also the fact that in 2005 there has been reform in the sex crime legislation, which made the legal definition of rape much wider than before.”

Apparently the Swedes also count sexual assault by individual act, even if the incidents were by the same perpetrator (ie if she was assaulted twice in one night, that’s two rapes….in the US we count people reporting they’ve been raped as one per person), though it’s unclear when that counting convention started.  There has also been a major push to attempt to increase reporting rates and increase police knowledge around the subject.

According to this Amnesty International report, there likely has been a real rise, though likely not as dramatically as the chart above would suggest.  Apparently the number of cases being brought to trial has risen, but the conviction level has stayed fairly steady.

Off topic, but the BBC article above also points out that Australia and Canada have the highest kidnapping rates in the world, in large part because they count custody dispute issues (where one parent takes the child during non custodial time) as kidnapping incidents.  Interesting stuff.

Poison: the preferred weapon of women, cravens and eunuchs

/ bs king / 4 Comments

A couple of weeks ago I was watching something – can’t remember what – and I heard someone casually mention that someone who had been poisoned was likely killed by a woman.  It’s a trope I’ve heard before (both Sherlock Holmes and Ned Stark both assert this), but for some reason I’d never questioned it.  Anyway, I put it on my mental list of “things to google” before promptly forgetting about it until I saw this Wired magazine post.

In it, Deborah Blum (author of the Poisoner’s Handbook) asserts that homicide by poison is much more likely to be committed by a man (60.5%) than by a woman (39.5%), and that therefore the idea that poison was a woman’s weapon was false.  Her numbers come from this report…snapshot here:
Weapon          Male         Women
Gun homicide   92.1%      7.9%
Arson              78.8          21.2
Poison             60.5          39.5
So women are less likely than men to poison, but they are better represented in that group than the two others in the report.  But I got curious….what does this mean in terms of absolute rates?  From the way the data’s presented, it does appear that if a woman murders, there’s a good chance she used poison….but is this true?
I took a look at the FBI crime database to see what the absolute numbers were.  The numbers above are for the years 1980-2008, and this report is for 2006-2010, but my guess is the order of magnitude holds.  
For 2006-2010, there were 47,856 gun homicides, 505 fire/arson, and 49 poisonings.  So despite the lower percentage, women are still almost 200 times more likely to kill using a firearm than poison*.  So basically, there is no base rate fallacy going on here.  If you hear someone was poisoned, it’s more likely a male did it (at least for the years listed), and if you hear a women killed someone, poison was not her most likely method.  
Of course poison may have fallen out of fashion a bit, so this trope could have been true in Sherlock Holmes’ day, and all bets are off in the fictional world of Game of Thrones.  In case you’re curious, the FBI does not appear to keep data on crimes committed by eunuchs, so I can’t verify any of that.
*This may be skewed, as it is far more likely that some poisonings got missed by coroners than gunshots…but I doubt the missed cases would make up much of the difference. 

Statistics 101

/ bs king / 1 Comment

The Assistant Village Idiot has up a short and sweet post on proof:

Whenever I come across the word proves in a news story or a comment section, I usually think “Here’s someone who didn’t take enough math courses.”

I feel the same way when some says “that’s statistics 101”.

Example: I was reading a story recently on a particular type of forensic testing that was coming under some question (I was a lab tech in a former life, these things interest me).  Anyway, the study author was quoted as saying that 15% of the samples they were able to test showed some contamination, with the caveat that only one third of the samples in storage were still testable and thus the percentage could be subject to change.

When I was reading the comments section, one of the commenters got quite irate that this was being presented as only a 15% potential error rate.  Since only a third of samples were tested, he claimed we should actually multiply by 3 to get the real error rate….45%.  That’s Statistics 101! 

Sadly this is a blog with lots of angry and under educated commenters*, so the next 3 follow up comments were all along the lines of “nice catch”.

Math, it’s how you know when people are lying to you (but only if you do it correctly).

*This is another “not going to link to it for fear of track-back vitriol” blog citation.  But if you’re curious, it’s a blog tackling the issue of false criminal accusations.  While it’s a real and important issue, it does attract a good number of irrational people who hate the world and leave comments expressing their feelings quite….disturbingly.  The guys who run it seem pretty fair though, and I like reading the forensics aren’t perfect stuff, CSI be damned.

Let’s get ready to rumble!

/ bs king / 7 Comments

For those of you who are more recent readers, you likely don’t know that in a former life I attempted to start a wrestling blog.  

I happen to enjoy professional wrestling pretty tremendously, and though I could never quite find the time/energy/consistent topical flow necessary to keep a whole blog going, I still get a kick out of some of the posts I did there.  
I bring this up because tonight is my absolute favorite wrestling pay-per-view of the year, the Royal Rumble, and I will be watching while catching up on some stats homework.  
The Royal Rumble is named after the marquee match of the show, a 30 (sometimes 40) man battle royal.  Participants enter the ring at various time intervals (normally every 60 – 120 seconds) and eliminate other contestants by throwing them over the top rope.  Last man standing wins.  
This setup lends itself to some of the best statistics in sports entertainment* as this webpage shows.  Entrant #27 is the most likely to win (4 times) followed by #24 (3 times) and #1 and #30 (2 winners each).  
Anyway, if you’re surprised by the fact that I like wrestling, and would like to know the statistical probability of me being a wrestling fan, I did a post on that here. If you’re feeling more political, here’s my take on which wrestler I think various Republican media figures would be if they had to be wrestlers.
And just for giggles, here’s the number one anti smoking PSA of all time:
Alright, less than 30 minutes til start time, I’ll let you know who wins.   
*That’s what you call a sport whose outcomes are all predetermined….not fake mind you, predetermined.

Five reasons to check the footnotes

/ bs king / 4 Comments

I was flipping through the Volokh Conspiracy yesterday when I stumbled upon an article that revisited an incident involving their contributor Jim Lindgren.  

Apparently about a decade ago, there was a book out that claimed that very few people in early America owned guns, and therefore the 2nd amendment couldn’t possibly have meant that individuals should have had the right to own personal firearms.  Upon closer examination however, most of his footnotes and sources were fabricated.  Lindgren was a co-author on the article that took the book author down and completely turned his point against him….all because they actually bothered to track down the small print.
Interesting stuff.
If you really disagree with something, it’s always worth checking out the footnotes for a few things:
  1. That the source cited actually exists
  2. That the source cited backs up the part of the sentence that really needs backing up.
  3. That the source cited actually backs up the thing it’s being used to back up, and doesn’t just reference it obliquely.
  4. That the source cited states the point as strongly as the article authors state it.
  5. That the reference isn’t so old as to be outdated, replaced, or from a paper that has been unreplicatable.
I’m not saying everything you disagree with can be undone using these, but it’s pretty amazing how many citations don’t pass these 5 tests.  

Spam notice

/ bs king / 1 Comment

Just wanted to let you all know that I just released quite few comments from the spam folder.  Not entirely sure what happened, but I was getting notification for comments that were subsequently not appearing on the blog, including some from regular readers (karrde in particular seemed to have a few routed that way).

Anyway, I found them all in the spam folder, and made sure they got posted.

This also explains why there are a few right answers for yesterday’s brain teaser…..Eric’s answer was one of the ones that got caught up in the filter, and thus wasn’t there when Geek Vader got it right, even though it’s time stamped several hours before.

I’ll be more vigilant about this in the future.

Wednesday Brain Teaser 1-23-13

/ bs king / 9 Comments

When I was younger I used to spend a pretty strange amount of time reading through brain teaser books. As such, I occasionally hear a brain teaser and know the answer without being able to remember how in the world you get there. That’s what happened with today’s teaser…. I heard it on the Skeptics Guide to the Universe podcast this morning, answered immediately, then spent the rest of the train ride working out why I was right.  I got there somewhere near Hyde Park.

Lets see how you do:  A jeweler has 9 pearls, all identical shape and feel.  He knows one weighs slightly more than the other 8, but all he has to measure with is a balance scale (one with two arms that compares weights to each other). What is the minimum number of times he needs to use the scale in order to figure out for certain which is the heavy pearl?