Why is the suicide rate so low at Foxconn.

Foxconn is a chinese company that you may have heard a lot about recently. There has, according to the BBC been "a string" of suicides there recently. According to the Times there is a "spiralling suicide crisis". Reuters says there has been "a spate of employee deaths".

10 people have committed suicide this year at a particular Foxconn factory in China (where a lot of parts for the iphone are made, apparently, which I think is why this is supposed to be a news story). The factory employs 300,000 people. I make that a suicide rate of (approximately) 7 people per 100,000 per year. The suicide rate in China is about 13 people per 100,000 per year. In other words, working for Foxconn cuts your chance of committing suicide almost in half!

How is it possible for every single person everywhere in the world's media to have gotten this story exactly backwards? There are a lot of very clever people suggesting "explanations" for the suicide rate at Foxconn. Unfortunately, they're trying to explain why it's so high, when they should be doing the opposite! This should lead us to seriously doubt these people's 'explanations' when the phenomena they're describing happen to be real. If your theory can explain anything, it has no explanatory power at all.

However, I think the main story here is how this became a story. What is going on? Why did someone decide to report the suicide rate at Foxconn instead of, say, the suicide rate among Tesco employees (another company which employs around 300,000 people, at least in the UK). And how did no-one notice that what they were reporting was in no way interesting.

Deadweight Loss

I'm currently reading Joel Waldfogel's book "Scroogenomics". In it, he expands on some of the points made in his classic research paper "The Deadweight Loss of Christmas". The basic premise is that people (especially distant relatives) don't know what you want as well as you do, so any gifts they buy you are less desirable to you than anything you could have bought yourself with the money. He has plenty of facts and figures to back up this eminently plausible theory, and I think a serious point to make.

However, he does have at least one minor slip. In describing what deadweight loss is, he says the following:

If a dollar disappears from my pocket and appears in yours, it's a loss to me, but it's not a deadweight loss to society. If you take my dollar and destroy it lightling your Cohiba, then it's a deadweight loss.

This isn't quite right, for the simple reason that dollars only have purely symbolic value. Burning a dollar bill is a genuine loss of $1 to the burner, but can't possibly make society as a whole worse off... you can't eat money. So who benefits? As the wikipedia article on burning money explains, everyone. Burning a banknote has a (very) slightly deflationary effect, so makes all of the money in everyone else's pocket worth slightly more.

I first encountered this idea in Steve Lansburg's excellent Armchair economist, and it seems to be pretty standard fayre in economics literature, so how did Waldfogel miss it? Or did he think that burning money seemed more wasteful than any other genuine example of a deadweight loss, so sacrificed accuracy for impact?

What exactly are confidence intervals?

So, a friend (who would probably prefer to remain nameless) is currently looking for jobs in finance, and sent me some sample interview questions that they like to ask people. One of them is the following:

A drug trial gives the result that the drug works better than the
placebo, with 95% confidence. What exactly does this statement mean? What further assumptions are needed to be able to deduce that the probability of the drug working is actually 95%?

Now, I wasn't entirely sure I knew the answer to this question. I thought I did, but despite having studied rather a lot of statistics, I don't think anyone ever actually told me what a 95% confidence interval was, so I did what everyone does faced with such a situation, and checked the wikipedia article. Wikipedia is singularly confusing on the matter, but it gives the answer as roughly the following.

Click here to read the rest

My Favourite Limericks (of which precisely one is actually a limerick)

Here's the actual limerick:

There was a young man from place B
Who satisfied Predicate P,
He performed action A,
In adjective way,
Resulting in Consequence C.

Click here to read the rest

Wikipedia can't add up

Have a look at this list of best-selling books. In particular, look at the Harry Potter books. There is one Harry Potter book in the list of best-selling books. Harry Potter and the Deathly Hallows, which has sold approximately 44 million copies. Harry Potter is also listed as the best-selling series of all time, with 400 million total copies sold.

In order to see why there is a problem with this, you just need to know that there are 7 Harry Potter books, and understand the Pigeonhole Principle.

There is a (sort of) explanation of the problem in the 'talk' page of the wikipedia article, but the inconsistency remains on the initial page. This actually highlights a fairly major problem with wikipedia's epistemology. The inconsistency has to remain because there's no 'credible source' for the figures for the other individual books. So even though everyone knows that the article is wrong, it can't be corrected because this would constitute 'original research'. Hmm... just how credible does a 'credible source' have to be? Will a mathematical proof do?

How to sample randomly

I was once told that 'duration of unemployment' figures were collected in the following way: people were telephoned at random during the day. If they answered, they were asked if they were unemployed. If they said yes, they were asked for how long they had been unemployed. Before you read any further, can you see what is horribly, horribly wrong with this method of data collection? (There are several things wrong with it, but one of them renders it entirely useless)

Before I give you the answer, a brief detour. When I was at school, I did a piece of statistics coursework (I think it was for GCSE's) in which I compared the average sentence length in a French text to an English text. I can't remember exactly which texts I chose, I think it was newspapers of 'equivalent' quality, but that's largely irrelevant. In order to estimate the average length of sentences in each text, I adopted the following method: pick a word uniformly at random from the text and count the number of words in the sentence containing it.

I had collected around 100 sentence lengths before I noticed the utter ridiculousness of this method. In case anyone hasn't spotted it yet, this 'random sampling' is guaranteed to massively overestimate the average sentence length in any given document, as the probability of any given sentence being chosen is in direct proportion to its length.

Consider the following passage:

"The quick brown fox jumps over the lazy dog whilst the five boxing wizards jump quickly over my lovely sphinx of quartz. Jesus wept"

 If we pick a few random words from this and compute the 'average' sentence length of the sentences that contain them, we're going to come up with something very close to 20 (if we pick every single word, we'll get 20.3333)  The actual average sentence length is 12.

Now, if you didn't immediately spot that this was the key problem with the method of collecting unemployment data I mentioned in the first paragraph (there are problems with telephone polls in general, of course, but they are essentially insignificant compared to the problem with the sampling method), this should make you worry about how easy it is to slip *exceedingly* dodgy statistics past people who aren't paying attention. I'll post a few examples of my favourite 'correlated for spurious reasons' statistics in another post later this week.

As an aside - if you do actually collect the data in the way suggested, you can presumably still get some information about the distribution you're studying - what's your best estimator for the mean? And what assumptions do you have to make about how the data are distributed?

Why don't we sample more?

Steve Landsburg recently blogged about a maths professor who weeds out 'unlucky' applicants by randomly rejecting half of the resumes he gets sent. Now, this is unusual, in that it is a random sampling method which significantly *reduces* the average quality of the applicant that gets hired.

There are a *lot* of situations in which random sampling would reduce workload whilst having no effect whatsoever on effectiveness. I'll start with one of the simplest and least controversial (and one that I have the most personal experience with). Students regularly submit 10 or more pieces of coursework for each course in a university semester. Every question is then marked, and the papers returned to the students. Assuming (which is probably not entirely accurate) that the courseworks are solely intended as a normative assessment of student performance, surely it would be massively more efficient to sample questions at random and mark those, rather than marking the entire paper. The expected mark for any given student is the same - only the variance goes up.

There are a few situations in which students suffer as a result of this. Say there's a pass mark of 40, and you have to pass every coursework, now someone who answers exactly 40% of the questions right in each coursework expects to fail (although they do expect to get an average mark of 40). Similarly, there are situations in which students benefit from this (pass mark of 40, answer exactly 39% of the questions correctly, you now have a non-zero chance of passing). On the whole, I would expect these things to cancel out, and that no one student knows their mark accurately enough to know whether they would benefit or lose out from this policy being enacted.

So why isn't this done more? I've heard from a few lecturers who've tried it, and it went down horribly with the students, who perceive it as 'unfair'. Apparently there were several comments along the lines of 'what if you only mark the questions I did badly?'. I guess this is some sort of loss aversion - it is quite obviously equally likely that we only mark the questions you did well!

Yvain has an article about a similar example from education - in which students are reluctant to guess answers to true/false questions with a penalty of 50% of a point for a wrong answer for some inexplicable reason. Again,  random sampling is a massive net win.

Another example is public transport. No-one every pays to get on the 25 bus. This is because it is extremely rare for anyone to check whether you've paid or not and the penalties just aren't high enough to make it worthwhile paying given how rare the checks are. There are two obvious solutions to this problem: you could either do twice as many checks (thus requiring you to hire twice as many people to do the checking, and inconvenience twice as many people whilst checking) or you could double the fine. I've no idea why they don't take the second option.

How about voting? Instead of counting all of the votes in a general election, why not shake the votes up in a big bowl and count, say, the first 10,000 for any given seat? I can't be bothered to crunch the numbers, but I'm pretty sure the probability of error would be down below 1% - and errors would only occur in seats which were closely contested - where errors are not so important anyway, as the people obviously don't have a clear preference between the candidates.

Most of the examples I can think of exploit the same principle as the public transport idea above - when committing some transgression, your expected utility is the utility of cheating minus the disutility of punishment times the chance of getting caught. Since it's expensive to increase the chance of getting caught, there are a lot of situations in which I think it would be a net win to decrease this and increase the size of the punishment. Why not check half as many tax returns and double the fine for misfiling? Have half as many speed cameras and double the fine for speeding (speed cameras aren't expensive, so this might not be a net win)?

There are dozens of examples - and I don't think that the people in charge have sat down and done the relevant calculation in all cases. Are people just afraid of randomness? Afraid of seeming 'arbitrary'? Afraid of letting people 'get away with' committing crimes - assuming the only legitimate purpose of the criminal justice system is deterrence, this shouldn't be an issue. Maybe there's legitimate concerns that a 'random sampling' approach to some of these problems would be more subject to corruption - but we can just check a few of the samplers at random, and have massive fines for people doing it corruptly!

The law of large numbers is a powerful and important mathematical theorem. Why don't we exploit it better?

Lucozade Sport Lite: the Low Energy Energy Drink!

I just saw an advert for Lucozade Sport Lite. Yes. Seriously. This is an energy drink which contains only 50 calories. But don't worry:

Lucozade Sport Lite contains electrolytes and fluid which help to keep you hydrated

Hmm... a fluid which helps keep you hydrated, I wonder what other things there might be that fit that description....

The website really is excellent though - it actually include a section entitled 'how do I use it?'. Erm... put it in your mouth and swallow?

Gladwell on probability

There's quite a nice list of random quotes from Malcolm Gladwell in an interview for this Sunday's Observer.

However, one of them seems to show some misunderstanding of probability:

History suggests that there is almost exactly a 50% chance that any piece of information a spy gives you is true. We would be as well off getting rid of the secret service and flipping coins.

Now if the first part of this sentence is true (which I have no reason to doubt) the second part most definitely does not follow. This is (tangentially) related to a discussion that's been going on at Peter Cameron's blog about probability. Unless spies only ever make statements about things where your prior was already 50%, a 50% accuracy rate could be incredibly useful.

Eg, let's say we're trying to find out where a particular terrorist group has their headquarters. To start with, our probabilities are essentially uniformly distributed across the whole of the world. Our spy comes up to us and says 'the HQ is at number 32 Barkston Gardens, Earl's Court, London'. This information is far from useless - in fact, if we have more than one spy coincide on the same piece of information then we're in business, and can find the location pretty quickly.

Of course, I think Gladwell's '50%' is actually just a proxy for 'exactly as true as you'd expect if they were generating their statements at random', but that's not *quite* the same thing

Whacky World Cup Formula: Germany will win

According to this article from the Telegraph (which, incidentally, is a carbon copy of the article in various other news sources from around the world - I assume it's lifted directly from a wire service, but who knows?) Germany are definitely going to win this year's world cup. How do we know this? Trigonometry!

The scientist has written a formula based on trigonometry which analyses all Germany's results from previous World Cups and predicts a winner for this year's tournament.

Having won the World Cup three times, in 1954, 1974 and 1990, Germany's average finishing place at previous tournaments is 3.7 and Prof Tolan says his formula shows this will be Germany's year to lift the trophy. 

Unfortunately, Metin Tolan doesn't show his work, so we can't see quite how he came up with this ridiculous conclusion. It's hard to see why any team with an average finish of 3.7 (whatever that means) would expect to finish in position 1. It's also really, really hard to see how a formula which appears to be essentially some sort of regression/reformulation of the law of averages could be 'based on trigonometry'.

We do get some indication of his track record:

Prof Tolan already predicted Germany would win the last World Cup, which they hosted in 2006,

So he may be displaying a tiny bit of overconfidence in his predicition to say:

"Nobody can beat us this year and you can already put the champagne on ice."

However, my favourite part of the article is the end, in which Tolan demonstrates that he doesn't understand basic game theory or probability, with regard to penalty shoot-outs:

"The weakest kicker should take the first penalty, then the second-weakest and so on," he said. "Then you have the greatest chance of scoring as many goals as possible."

Now this is quite clearly the exact opposite of the truth. I can't even be bothered to crunch any numbers, because it only takes a few seconds of thought to see that if one side follows the good professor's advice whilst the other uses the more sensible plan of doing the exact opposite, the latter side will win before things have even gotten started - the better players on Tolan's team literally won't get a kick.

I'm not quite sure why people who seem to have perfectly respectable research careers get inolved in this sort of thing. I would suggest it was for the money, but he did the same thing 4 years ago... is it just that some people can't resist having their name in the paper?

What You Can't Say in Harvard

Paul Graham  is one of my new favourite authors, and What You Can't Say is one of my favourite among his essays*. He discusses the idea that there are almost certainly several things which happen to be true, but which we can't refer to in polite society. I want to write about recent example of the sort of topic that appears to be taboo in modern western academia. To quote Graham:

What can't we say? One way to find these ideas is simply to look at things people do say, and get in trouble for.

That seems like a perfect introduction for this story which, if it weren't true, would strike me as utterly implausible. A Harvard Law student wrote a perfectly reasonable email to a friend, about six months ago, in which she stated that she could not "absolutely rule out the possibility that African Americans are, on average, genetically predisposed to be less intelligent." She was roundly condemned by the Harvard Black Law Students Association, and the Dean of Harvard Law.

Now... I'm going to take a risk and say that I can't 'absolutely rule out' the possibility that African Americans are genetically predisposed to be less intelligent than non-African Americans either (although Stephanie Grace is much better informed on this topic than I am).  Presumably, though, there is no a priori reason to assume that the intelligence of the two groups is the same - African Americans, for example, are clearly genetically predisposed to run faster (I'll give anyone 50-1 on a white man winning Olympic gold before 2020!), why should intelligence have a smaller genetic component than running speed? 

What is much more disturbing is the reaction from the Dean of Harvard Law:

I am writing this morning to address an email message in which one of our students suggested that black people are genetically inferior to white people.

Firstly, she doesn't appear to have read the email in question. The student actually suggested that she wasn't convinced either way by the evidence. Secondly, and more worryingly, the Dean of Harvard Law is equating 'more intelligent' with 'superior'. She seems to be worryingly close to implying that *if* it turned out that black people were genetically predisposed to be less intelligent than white people (which is it at least a logical possibility) then this would somehow vindicate racism. And even closer to implying that stupid people (who definitely *do* exist) are somehow less worthy than intelligent people.

It is very dangerous to attach moral weight to issues of scientific fact. There is always the danger that you might be wrong.This is an issue that Pinker comes back to again and again - there are a list of supposedly morally charged issues of scientific fact in the preface to "What's your Dangerous Idea?" If you are wrong about any of these issues, are you willing to bite the moral bullet?

Does the Dean of Harvard Law really believe that racism is wrong because race differences in intelligence are negligible? Or does she believe it is wrong because a person's worth isn't related to their intelligence, or any other trait they may have, but a fundamental part of being human?

This might just turn out to be another one of those cases where reality is the Least Convenient Possible World. I'm willing to countenance that possibility and still condemn racism. So is Stephanie Grace (I'm charitably assuming her apology is a political necessity, rather than a genuine retraction of her commitment to scientific integrity). Is Dean Minow? 

* Although Why Nerds are Unpopular has some important insights, and contains several ideas which make you go 'I wish I'd thought of that first' - or, more precisely, that 'I wish I'd managed to formulate so clearly first': 

It's important for nerds to realize, too, that school is not life. School is a strange, artificial thing, half sterile and half feral. It's all-encompassing, like life, but it isn't the real thing. It's only temporary, and if you look, you can see beyond it even while you're still in it.