Ben Goldacre masterfully dismantles the dodgy science behind some of the great drug trials, court cases and missed opportunities of our time, but he also goes further: out of the bullshit, he shows us the fascinating story of how we know what we know, and gives us the tools to uncover bad science for ourselves.

I see from today’s Volkskrant that Lucia de Berk has at last won the right to a retrial of her case. De Berk is a nurse who was convicted in 2003 for the supposed four murders and three attempted murders of patients in her care.

The history of the case makes chilling reading, not because of anything that de Berk may have done, but because of the web of statistical “proof” that the prosecution used to put her behind bars. It is perfectly clear that the statistical evidence was deeply flawed from the start, but here we are in 2008, and she has spent almost six years in jail for “crimes” that never existed in the first case.

The judgement against her was based largely on the claim (from the prosecution’s statistician) that the chances of so many people dying on the wards where she was on shift were “one in 342 million to one against”. But, as Ben Goldacre makes clear in his excellent book Bad Science, the fundamental flaw about this claim is twofold. First, the data was selected to make the hypothesis, and then the prosecution’s statistician made a simple, rudimentary error: he combined individual statistical tests by multiplying p-values (the mathematical description of chance, or statistical significance). As Goldacre points out in respect of the first part of the claim:

A huge amount of corollary statistical information was almost completely ignored. In the three years before Lucia worked on the ward in question, there were seven deaths. In the three years that she did work on the ward, there were six deaths. Here’s a thought: it seems odd that the death rate should go down on a ward at the precise moment that a serial killer – on a killing spree – arrives. If Lucia killed them all, then there must have been no natural deaths on that ward at all in the whole of the three years that she worked there.

And in respect of the second flaw, Goldacre points out:

If you multiply p-values together, then harmless and probable incidents rapidly appear vanishingly unlikely. Let’s say you worked in twenty hospitals, each with a harmless incident pattern: say p=0.5. If you multiply those harmless p-values, of entirely chance findings, you end up with a final p-value of 0.5 to the power of twenty, which is p < 0.000001, which is extremely, very, highly, statistically significant. With this mathematical error, by his reasoning, if you change hospitals a lot, you automatically become a suspect. Have you worked in twenty hospitals? For God’s sake don’t tell the Dutch police if you have.

It’s a very cautionary tale of statistics gone horribly wrong, and very reminiscent of the Sally Clark case in the UK (which Goldacre also dissects). Clark was put on trial in 1999, and convicted, for murdering her two babies. At the trial, child expert Professor Sir Roy Meadows stated that the chance of two children in the same family dying of Sudden Infant Death Syndrome (SIDS) was “one in seventy-three million”. It was another case of statistics wielded in error, and Clark spent three years in jail (where she was targeted by other prisoners as a supposed baby-murderer) before her conviction was quashed by the Court of Appeal. She emerged a broken woman and died in March 2007. I fervently hope that that will not be the fate of Lucia de Berk.

So, sure, Barack Obama might be every bit as much a slick piece of imageering as Sarah Palin. The difference is in what the image represents. The Obama image represents tolerance, intelligence, education, patience with the notion of compromise and negotiation, and a willingness to stare ugly facts right in the face, all qualities we're actually going to need in government if we're going to get out of this huge mess we're in.

 

Here's what Sarah Palin represents: being a fat fucking pig who pins "Country First" buttons on his man titties and chants "U-S-A! U-S-A!" at the top of his lungs while his kids live off credit cards and Saudis buy up all the mortgages in Kansas.

 

The truly disgusting thing about Sarah Palin isn't that she's totally unqualified, or a religious zealot, or married to a secessionist, or unable to educate her own daughter about sex, or a fake conservative who raised taxes and horked up earmark millions every chance she got. No, the most disgusting thing about her is what she says about us: that you can ram us in the ass for eight solid years, and we'll not only thank you for your trouble, we'll sign you up for eight more years, if only you promise to stroke us in the right spot for a few hours around election time.  

Religion, uncontaminated by power, can be the source of a great deal of private solace, artistic inspiration, and moral wisdom. But when it gets its hands on the levers of political or social authority, it goes rotten very quickly indeed. The rank stench of oppression wafts from every authoritarian church, chapel, temple, mosque, or synagogue – from every place of worship where the priests have the power to meddle in the social and intellectual lives of their flocks, from every presidential palace or prime ministerial office where civil leaders have to pander to religious ones.

 

My basic objection to religion is not that it isn't true; I like plenty of things that aren't true. It's that religion grants its adherents malign, intoxicating and morally corrosive sensations. Destroying intellectual freedom is always evil, but only religion makes doing evil feel quite so good.

My understanding of economics is pretty basic. I can understand, and try to follow, the financial principles espoused by Mr. Micawber in Dickens' David Copperfield:

"Annual income twenty pounds, annual expenditure nineteen pounds nineteen and six, result happiness. Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery."

I really think that that is all one needs to know, anything more, it seems to me, gets closer to religion and blind faith. So I can't say that I'm much surprised by the current crisis in the financial markets.

In an attempt to understand the causes, I've been reading two analyses of the events. The first is an article in Edge written by Nassim Nicholas Taleb called The Fourth Quadrant. I will forgive him the air of "I told you so" that permeates the article, he's somewhat justified in having it. After all, it was he who wrote back in 2006:

The government-sponsored institution Fannie Mae, when I look at its risks, seems to be sitting on a barrel of dynamite, vulnerable to the slightest hiccup. But not to worry: their large staff of scientists deemed these events "unlikely."

However, his analysis goes into dizzying detail about statistical theory and how people misinterpret statistics and probabilities. I'll probably need to read the article several times before I can come closer to understanding what he is saying. But even after all that, I can't help feeling that he's shining the light of reason in the wrong place. That was brought home to me when I read John Carter Wood's article over at Obscene Desserts. Like me, John finds the workings of high finance rather mysterious. He quotes from an article in the New York Times:

The mortgages, with an average size of about $450,000, were Alt-A loans — the kind often referred to as liar loans, because lenders made them without the usual documentation to verify borrowers’ incomes or savings.

And then, without the benefit of graphs or statistics, John gets to the root of the matter:

Let's just pause here for a brief moment. Just for a measly few seconds.

Please just consider that last sentence, the one in which it is pointed out that lenders gave people mortgages worth an average of nearly a half-million dollars without even checking how much they earned or how much money they had?

Is this for real?

Because if it is, I can only say: What--please pardon my French (you know, I've been spending some time there)--the fuck?!

It seems to me to be all too easy to label the shenanigans on Wall Street and in other financial centres as “a crisis too complex for easy fixes”. John’s put his finger on it, the world’s economy is seemingly run by a bunch of greedy wankers. Micawber’s principles should never have been forgotten.

Metadata is “data about data”. While the term is used to cover a wide variety of fields, the one that most interests me is metadata used in digital images.

The most prevalent standard for digital image metadata is Exif, implemented in every digital camera. It’s metadata focused on technical information about the picture itself: camera shutter speed, exposure, etc. The standards of IPTC-IIM and XMP are focused on metadata about the subject of the photograph: topic, location, photographer, etc. IPTC-IIM is an earlier standard that has been in use by the press agencies since the 1980s - so there's a large installed base. However, it's not easily extensible. Enter XMP, which is XML-based, first created by Adobe Systems Inc. in 2001. In 2004, Adobe got together with the IPTC organisation and others to create the “IPTC Core” standard, an XML-based successor to IPTC-IIM.

Of course, as the old joke has it, “the great thing about standards is that there are so many to choose from”. What is worse, is that different manufacturers (of cameras, application software, operating system software) interpret the standards in subtly different ways so that the resulting implementations are often incompatible with each other.

This got to such a stage that in 2006, Microsoft proposed the setting up of a cross-industry working group to deal with the matter. As a result, the Metadata Working Group was created in 2007 by five founding members: Apple, Adobe, Canon, Microsoft and Nokia. In 2008, Sony joined the party. The goals are:

• Preservation and seamless interoperability of digital image metadata.
• Interoperability and availability to all applications, devices, and services.

Not before time, I hear you cry. Certainly I get very frustrated by the current lack of interoperability that exists.

Yesterday, at the bi-annual Photokina exhibition, the Metadata Working Group released version 1.0 of their Guidelines for Handling Image Metadata. In the Group’s own words, “this document describes how best to use existing standards such as Exif, IPTC, and XMP to address the key organizational metadata questions that most consumers have”.

I’ve only had time to quickly glance through the document, but it does seem to be a good effort to attempt to herd the bunch of cats that are software developers – many of them in their own companies.

It was with some irony that I noted this paragraph (page 32):

Hierarchical keywords are not covered. However it's well understood that this is an important use case even in the context of the consumer and will be added to future versions of this document. There are existing solutions available e.g. Adobe Bridge, Adobe Lightroom as well as Microsoft Expression Media and Windows Live Photo Gallery that have introduced hierarchical keyword workflows specific to their needs.

Irony, because it’s the way that Microsoft’s Windows Live Photo Gallery has currently implemented hierarchical keywords that is causing me some interoperability problems and frustration at the moment.

Oh well, back to herding cats…