Why Dont You Blog?

Well, it seems the UK tabloid press is continuing its efforts to make “middle England” terrified of shadows without any real basis. Yesterday’s Mail on Sunday has ensured that its readership have “evidence” that Britain is descending into anarchy and the police crime statistics (as well as the victim reported data in the British Crime Survey) is just nonsense.

In a nutshell, the article is about Ms Sarah Schaefer (senior adviser to Foreign Secretary David Miliband) who was “carjacked” in a posh London street last Tuesday. She was driving along the street, when a “thug” jumped out in front of her, forced her to stop and jumped in the passenger seat. Ms Schaefer fled the car and threw the keys away (obviously the car was more important to her than any mere prevention of harm). The unnamed “thug” found the keys, got in the car and fled with it (later crashing).

Now all in all, this is a reasonably traumatic experience and it is sad that Ms Schaefer underwent it. However as far as the Daily Mail is concerned this is proof that the UK is in a grip of unprecedented levels of crime - despite any claims to the contrary by the police or government. Very early in the (erm) article, the breathless “journalists” write:

The ordeal of Sarah Schaefer is a major setback to Labour’s rubbishing of Conservative claims that the rise in violent crime has led to “anarchy in the UK”.

I know I can be slow on the uptake but I don’t get this. How does ONE crime support the Conservative’s claims? Is there some mystic aura about Ms Schaefer which means she can only become the victim of crime when 75% of the population has been? She is one person. Nothing in the article gives any indication as to the true rates of this type of crime (check BCS if you are that bored) but it has this bit or terror inducement:

The attack on Ms Schaefer is a stark reminder that crimes such as carjacking, once associated only with ghettos in the US and South Africa, are now commonplace here – and can occur in neighbourhoods popular with the middle classes.

This is mind boggling. Carjacking is not commonplace on the mainland UK. For those unfortunate enough to live in Northern Ireland, however, carjacking is more common and has been for 2o years. The sad part is the Daily Mail (and its readership) would never want to let facts or statistics get in the way of a good bit of fear.

Just in case the (insane?) middle England readership of the Mail missed the point they were trying to be given, the article finishes with:

Ms Schaefer is just the latest highprofile person to fall victim to rising crime.

Muggers stole a mobile phone from Richard Madeley and Judy Finnigan’s daughter Chloe, 19, while she was making a call in Notting Hill’s fashionable Portobello Road in March.

Chelsea and England footballer Frank Lampard’s £8million West London home was burgled in May 2005 as the star and his girlfriend Elen Rives slept upstairs.

And high-profile divorcee Beverley Charman, 54 – awarded a £48million payout – was tied up at her Kent home and robbed of jewellery worth £300,000 in March.

This is more of the odd way the media seem to blow the lives of the rich and famous out of all proportions. The claim that this is the result of “rising crime” is more than misleading, it has no basis in fact and it certainly is not supported by anything in the newspaper. There are thousands of “rich and famous” people who live in the UK. If you include “high profile” then we could have in the region of 100,000 people to consider. This newspaper article identifies FOUR who have been the victim of crime and seems to cover the period of May 2005 - Sep 2007.

This makes the rate of crime around 1.3 per 100,000 people per year - if this is “rising” how low was it in the past? If this is really representative of the nation (as the article seems to imply) then we have a crime rate of 871 crimes throughout the UK per year. Blimey. What a safe nation.

The only way I can see that this article tries to show “rising” crime is that there seems to have only been 1 high profile crime between May 2005 and March 2007, but since March this year there have been three. Even then it is farcical.

Sometimes I really do wonder what goes on in the minds of people who read this sort of drivel and believe it (check the comments out if you want a laugh). Most of the Mail readers I have met in real life actually hold to the ideas the paper puts forwards (much to my frustration), most are from fairly affluent backgrounds and most have never been the victim of any crime in their life. Despite this all talk about how “bad” things are, how children are unruly, how crime is out of control and how someone they know, knows some one who has been burgled. It almost makes me want to cry.

[tags]Daily Mail, Mail on Sunday, Rant, Society, Law, idiots, Idiocy, Bad Journalism, Crime, Rich People, Famous People, Crime Rates, British Crime Survey, Statistics, Bad Statistics, Carjacking, Sarah Schaefer, Anarchy, Britain, UK, Criminals[/tags]

Popularity: 56% [?]

Popularity: 56% [?]

In today’s Guardian newspaper (and online and here) there is an article explaining how the fears and worries of the “electrosensitive” woo-mongers is unfounded.

Sadly, the Guardian’s “news” editors have chosen to go with the headline:

Research fails to detect short-term harm from mobile phone masts

Now, it may just be my pedantry, but surely that strongly implies there is a short term harm and the researchers simply failed to detect it? The second link above is better and carries the tag line:

Yet another study shows no link between mobile phone radiation and ill health

Which pretty much captures the repetitiveness of this as a research result. The overwhelming weight of science shows there is no evidence of any short term harmful effect from communications masts and the only proven long term risk is from the most popular source of electromagnetic energy itself - the Sun.

In a nutshell, this seems like a well designed study which, like all the others, has resulted in no evidence that people who claim to be sensitive to electromagnetic radiation actually are - this is even something I have mentioned in the past. Repeated tests have shown that if you get an “electrosenstive” and tell them there is a transmitter near by, they evince the effects they claim are caused by “EM.” If they dont know the transmitter is near by, they don’t have the effects. In my unsympathetic, un-medical opinion this is pretty good proof it is all in their mind - for various reasons they are completely making it up. Part of me concedes the symptoms may be real, but it is only a small part of me. Either way, targeting phone masts as the culprit is doing no one any favours. As the Guardian comment on the topic finishes:

What sufferers experience is real and in many cases very unpleasant. But in the light of this evidence we can be pretty certain that phone masts do not cause short term health problems for the vast majority of people. Electrosensitive support groups should recognise this and begin to look harder for other causes of the condition.

Well said. Stop fighting a bogeyman and find the real cause - if there are real symptoms.

As always, there are those who are so wedded to a concept that no matter how much evidence to the contrary is presented, they will refuse to accept it. Sounds a bit religious to me, but never mind. The wonderfully named “Mast Sanity” website is a cited opponent of the recent study, and shows many of the traits you would normally associate with creationists trying to debunk evolution.

Unsurprisingly, Mast Sanity is a screaming example of bad science and a place where spurious arguments are used to dispel the results of the most recent study — I assume similar tactics were used on older studies, I didn’t look into the site that much, what I did read seemed like a check list of logical fallacies and debate-scoring tactics rather than anything reasoned. Some examples include:

We question why psychologrists are doing this research at all since physical changes to the skin and heart rates have been found in other research. Presumably the psychologists ‘believe’ this is all in the mind and this is what they set out to ‘prove’.

Yeah, and when you read the research notes it shows the psychologists set out to measure the physical responses. This smacks of a combination of appeal to ridicule and the laypersons perception that educational disciplines exist in complete isolation of each other. If the researchers had set out to prove the Electrosensitivity was in the mind, this would be obvious from the experimental design, not from what discipline the people who run the experiment come from.

Their conclusion was made possible by eliminating 12 of the most sensitive electrosentive volunteers who had become too ill to continue the study. Even a child can see that by eliminating 12 of the original 56 electrosensitive volunteers - over 20% of the group - that the study integrity has been completely breached.

Wow. First off the 12 people withdrew themselves, they were not eliminated to make the experiment possible. If the other 44 “electrosensitives” were actually electro sensitive, then what would the loss of those 12 change? As for the great “even a child” comment — well really. I have not met many children who can do the statistical analysis required to account for the changed sample sizes, but most would probably make a random assumption as to the status of the experiment. Does that mean they would be correct? Critically, the “study integrity” has certainly not been completely breached, it just gives a larger error bar to the findings.

There is more bad statistics with this bit of meandering nonsense:

One participant in the study questions Professor Fox’s assertion that only four people got all six test correct. He said “I got five [out of six] as during the first three five minute tests on session one, I stated ‘not sure’ after the first five minutes, which was marked as NO, but on session two, three and four I got it 100% right and actually identified the type of signal, so are the Essex [study] numbers meaningful?“

I will confess to not really understanding what this is trying to say. One person thinks that more (or less) than four people got all six tests “correct” because he got five out of six in one of them. Blimey. The whole experiment must be flawed then… I would really appreciate it if someone could explain what the above means to me — I must be having a bad understanding day today. Talking about a previous study, quoted by the BBC, Mast Sanity continues:

… We don’t think Dr. Rubin [author of previous study] is qualified to comment on the Essex study as he didn’t even use a shielded room for his own experiments at King’s College and the so called ’sham’ (zero) exposure was not a zero signal as people have been led to believe.

What makes me laugh about this, is the “pro-sensitives” leap on the shielding issue, and largely it is a cornerstone of their defence against the real science. In a nutshell, it explains why the “sensitives” report effects when no mast is transmitting, but they are led to believe it is. The problem with this is that when the “sensitives” believe the mast is off, they report no symptoms. Is the shielding belief-powered?

With no signs of irony whatsoever, Mast Sanity finishes its tirade with this wonderful bit of woo-spin:

Mast Sanity Spokesperson Yasmin Skelt says “All in all the Media release of this study has been an exercise in spin and propaganda and a poor one for science.”

“It is the long term health effects where people are forced to live near real Mobile Phone Masts that count and this study in no way covers those.“

Great isn’t it? They refer to themselves in the third person and claim the science is spin and their spin is science. New Labour must love the world they have created.

The study was solid science. It certainly was not a perfect experiment, but few ever are. The conclusions drawn are sound and the reasoning is valid. The Woo-Monger reactions have been an exercise in spin and bad-logic, rarely coming close enough to science to be thought of as bad science. The study was very upfront — as have been the media reports — that this didn’t look at long term effects. Sadly, spinning the goal posts to a new location does not invalidate the research — not that the woo crowd have ever worried about that.

Asking if there are long term health effects is a good question, and an area where the research is sketchier which results in less certainty over the answers. That said, the common cries of the “electrosensitives” is that they suffer short term effects (which is why people buy “shielded curtains” and the like) and on this, it is quite probable that they are wrong. Redefining the criteria each time one is falsified is typical of another group who hold to nonsensical beliefs in the face of all evidence. Will Electrosensitivity become the Woo of the Gaps?

[tags]Media, News, EM,Woo, Science, Bad Science, Statistics, Bad Statistics, Electromagnetism, Guardian, Electrosensitivity, Nonsense, Society, Belief, Research, Experiment, Evidence, Logical Fallacy, Spin[/tags]

Popularity: 42% [?]

Popularity: 42% [?]

Research from the Food Standards Agency was reported as showing that the poor do not have worse diets than the rest of the population. I am all for truths that fly in the face of “common sense” but I am finding this quite hard to swallow.

As soon as you look at the specifics, this whole argument starts to fall apart for me.

The Food Standards Agency found that contrary to popular belief, nutrition, access to food and cooking skills are not much different in poorer families.

• Nutrition. Given that so much of nutritional science is founded on guesswork and can often barely be distinguished from the Gillian McKeith schools of science, I’m not going to do this one to death, except to say that the points that they notice any difference in - such as consumption of fruit and vegetables - are the very things the nutritionists keep saying are important for our health
• “Cooking skills?” Why would anyone asume that poor people are less able to cook. Well, it seems that the survey does suggest that the British poor are indeed too stupid to know how to eat food, apparently unlike poor people in the rest of the world.
  

  Men and women with a lower level of educational achievement tended to have a ‘less healthy’ diet than men and women with more education. Men and women with less education ate fewer vegetables and more chips, fried and roast potatoes. Less educated women also consumed less fruit and fruit juice.

  If educational level has any correlation with income (as we are told by other parts of government), doesn’t this suggest that the poor do have a worse diet? So it might be poverty rather than lack of education that leads to the duff feeding?

• “Access to food? ” What on earth does that mean? It appears from the FSA website that it means where we shop and how we get it home.
  

  About 80% of this group did their main shopping at a large supermarket. About 50% had access to a private car for shopping

  Hmm - car? Not markedly poor then, you would think.

But then:

Mean weekly spending on food and drink (including eating out, but excluding alcoholic drinks) was just under £30 for one-adult households, just over £50 for households containing two or more adults, £55–£65 for households with one adult and one or more children, and £80–£90 for households with two or more adults and one or more children.

(You have feel particularly sorry for the adults in a 2 or more adult household, scraping by on £25 or less per person.)

Popularity: 31% [?]

Popularity: 31% [?]

It seems this is the week for nonsense “science” being thrown about by people who really should know better. This latest instalment may not be bad science, there are lots of fallacies which may well apply, but I will leave that up to you to judge.

Here in the sunny green and pleasant land of the UK, the TV and Radio were carrying a news bulletin, which has been picked up in the print press today, which explained that a Charity (Alcohol Concern) was calling for the Government to ban children under the age of 15 drinking alcohol at home. Seriously. Alcohol Concern are concerned [puns always intended] that a Government report shows the number of 11 - 13 year olds who “binge drink” has increased dramatically (I do not know what the figures for this are, sorry).

Depending on which news / radio station you caught this on, the feedback was mixed. In some of the “older listener” channels, there was applause at such good suggestions and heartfelt condemnation of “today’s youth” who are all alcoholic rebels, unlike any other time in the past… On the “younger listener” stations this was met with outrage and shock anyone would be daft enough to suggest it.

Popularity: 39% [?]

Popularity: 39% [?]

At the risk of turning this into a new topic which is hounded to death on the blog, I found some more interesting comments about carrying guns in public - nothing new, they are just more recent ones on the More Guns, More Homicide post.

Previously I pointed out that, as a Brit, I found it odd that the desire to carry a gun while you go about your daily life is so strong in some Americans and that oddness remains. I still find it strange beyond belief that some one in a civilised western democracy can feel so “unsafe” they need to be armed when ever they are in public.

Now, this self defence argument for “packing heat” makes me wonder a bit more. For example, on the comments “ben” was asked the following question by SG (read original):

let me get this right Ben. Someone walks up to you in the street and sticks a knife in your face, says “gimme your wallet”, and you think you can draw your gun and threaten and/or shoot them before they stab you? Is this how the self-defence argument works? Or does it work by you pulling your gun before they pull the knife, i.e. shooting them if they look threatening?

While this seems like a reasonable method of explaining the self defence argument for carrying a gun the responses it drew included this from MarkP (”I’m an actuary, I own a gun but don’t carry it, and have no particular love for them”) (read original)

No, SG, he thinks if he pulls his gun, the knifewielder will see the gun and will flee. And he is substantially correct. People wield weapons mostly for bluff. If the knifewielder had wanted to just have a fight, he would have stabbed Ben without asking for the wallet.

The only thing funnier than paranoid gun nuts protecting their phallic symbols are gun-phobes revealing that they cannot think logically about guns for one second, and know nothing about the real world. I suggest this be alleviated by talking about “weapons” rather than guns, since guns are only the most effective weapons at killing, but certainly not the only ones. Just ask the people in Rwanda.

While on the surface this seems like a logical line of reasoning, it suffers from a logical fallacy, having said that his attempt to broaden the debate is worthwhile, but I suspect it is futile.

Now as I see it the fallacy is that he is assuming the gun wielder is not bluffing but the robber is. If the statement “people wield weapons mostly for bluff” is true, then it must also apply to the gun owner and therefore drawing the weapon does not carry any reason to assume the robber will run.
When the gun is drawn the situation escalates. The robber will suffer from an adrenaline rush and may well decide that running will result in being shot in the back and attacking the gun owner is the only option. On a purely technical point, if the knife wielder is close enough to be a real threat anyway, the hand gun is probably useless unless it is already drawn.

There is more though. Ben replied to the question with: (read original)

It works like this: If I thought my life was in danger, then I’d draw and shoot if necessary. If not, then he can have my wallet, car and any other inanimate object he likes.

It is an interesting conundrum he presents. Some one is threatening you with a knife and you may not consider your life was in danger? If there is no threat, he will surrender his objects but if there is a threat he will try to fight. Very unusual and difficult to imagine how it can work in practice.

There are a multitude of arguments for, and against, gun ownership. My personal thoughts about the escalation of violence may well suffer from the slippery slope fallacy but I doubt it. When one person is armed and the other isn’t deaths may occur. If both are armed deaths may still occur and may well be more likely. Is the belief the death may not be the victim sufficient grounds for people to carry guns? Would you rather be punched or shot as the result of an argument with another car driver? Would you carry a gun in case the other driver came up to you and started shouting? At what point would you draw the gun?

As soldiers are taught, once the weapon is brought into view the whole situation changes. If your opponent does not back down immediately you pretty much have to kill them. The more people carry guns, the more likely an otherwise heated situation will turn violent. Is this grounds for banning firearms in the US? I don’t think so, but then I don’t live there.

Popularity: 26% [?]

Popularity: 26% [?]

Now, it seems I was a touch hasty with my last post on gun crime and statistics. If I had spent more time reading the comments on the Deltoid post even more comedy bad-statistical goodness would have come to light.

As someone living outside the US, I have no real concept of how this is an issue which creates such a furore. I am dimly aware of the constitutional issues, but as far as I can tell the right to bear arms was there because the founding fathers didn’t believe in a standing army. I don’t think they envisaged a future where both would happen, but who am I to know.

The main pro-gun comments seem to come from a guy called “ben” who proudly proclaims “I pack heat” and while superficially sound, they don’t seem to survive detailed analysis. For example, the numerous studies which have linked gun ownership to homicide are disproven by a the apparent fact that the homicide rate in gun-carrying Seattle is lower than in gun-banned Vancouver. Interesting. I think it is important to highlight that the debate is about “homicide” not “gun crime” which is a mistake I have made in the past.

While it is entirely possible that the apparently obvious more guns = more crime assumption is wrong, this leads to more questions which could have interesting conclusions. If more guns != more crime, then why does the nation with the highest incidence of gun ownership have more crimes per person than other nations? Are God-fearing Americans simply more criminally minded than (say) Canadians?

Popularity: 25% [?]

Popularity: 25% [?]

Well, as the month draws to a close a new copy of .net magazine appears. Always good for some ranting and this month carries on the traditions. I haven’t had the magazine for long so this is an “early stage” rant and that should be borne in mind. Remember, despite my misgivings about some of the crap they vomit out publish I am still a subscriber so it cant be all that bad!

Things do get off to a bad start this month though. Page 11 (first page of content) is where the hits begin. Now I am aware that as journalists the “reporters” for .net should be excused somewhat when it comes to understanding the mechanisms of surveys but even so…

Under the headline “No site, no sales” they have a three column article about how recent “research” shows 85% of people polled would have doubts about buying from a company that didn’t have a website. It goes on to produce dire proclamations backed up with “hard figures” (for example: “67% of small businesses believe it would take ten times longer to create a site than the average” - what does that even mean???) and finishes with the amazing proclamation that “a shop should sell stuff, a club should have membership info and a hotel should have online booking.” Fantastic.

Now the problem with this: The survey was comissioned by 1&1 and surveyed 1848 people. The number of people is acceptable but very low to make a comparison nationally. The big warning sign is the fact the survey was comissioned by a web host which sells online site creators and small business tools. Without going into this too much, from what I can gather the survey was carried out online which increases the disparity.

It seems a reasonable assumption that people confident enough about the web to take part in these sort of online surveys (lightspeed is a good example) would also have a higher threshold for requiring a shop to have an online presence. If I carried out this survey in the local villages where I live, I very much doubt if 10% of people would expect a shop to have a website before they would buy. You dont go online to check out your local newsagent before you buy the paper for example.

The article appears to imply that for small businesses to succeed they need a website. This, while good for business, is not really true. Most small businesses are aimed at selling goods to the local community, and in this situation the website is pointless. No one goes on line to check if the shop 200m away has a website before they buy. I agree that any business wishing to trade on a larger scale should have a website, but even then it is hard to think that 85% of their customers require one.

Ask yourself, when was the last time you saw an offline advert for a company and checked to see if they had a website before you bought. I have never done it. I have checked websites of online companies (eBuyer for a recent example), but they are online so of course they have a website.

To add scorn to their shoddy standards, in the sidebar of the article they “Name and shame” three sites which have “dismal” websites. Apparently SiteMorse looked at the websites for the FTSE100 companies and graded them. As always, Tesco.com gets slated - “zero for functionality” - yet even in the article it says they get hundreds of thousands of online customers. Oddly, the disparity of this escapes the .net journalists.

Instead of slating the site - visions of over paid designers sitting around in berets tutting about the site spring to mind - surely this implies the industry needs to overhaul its “testing” procedures (if there are any… I suspect it is just on a whim). Saying “bad design costs customers” seems true and is logically sound - however then saying the top online sellers have bad design lessens the point drastically. Tescos has an excellent website which hoards of people use for online shopping. I have used it and like it. I find it very functional and easy to use. What are the testers criteria if this real world example of a success is graded a failure?

Can anyone tell me?

Popularity: 27% [?]

Popularity: 27% [?]

Sorry I couldn’t resist that. As someone who actually works with statistics, I can confirm almost any statistics you examine in detail turn out to be based on spurious data.

The Bad Science article was particularly effective in uncovering the mythical nature of the doubling (1.4 rounded down to 1 and 1.9 rounded up to 2). Ben Goldacre even managed to explain the impact of clustering in comprehensible terms. Wow.

Rounding formed the basis of the oldest forms of computer crime (Round the results of millions of money calulations down and you can amass and, obviously, remove the invisible money that results. 100,000,000 times .5p is half a million pounds. Nowadays, the banks are wise to it of course - they keep the invisible money themselves)

Statistics are used to support almost any argument. This is not a fault in the statistical procedures. (I am prepared to take their value as given, insofar as I will never understand half of them.) The problem comes from the way statistics are embedded in social relations. Pure research for its own sake is rare. Who would pay for it? So any published statistics are created for some purpose - to direct the use of resources, to evaluate the success of a policy, and so on.

The “so on” includes influencing public opinion. The media are not interested in publishing dull statistics but they are very keen on inflaming public opinion. We have an infinte appetite for hearing that the young people are behaing worse than in the past, so how could a newspaper resist a headline grabbing statistic that involves youth, drugs and things getting worse?

Many thanks to Ben Goldacre for paying attention to a use of statistics and for having the wit to ask the rare questions - what are these numbers? what do they mean? who collected them?

Popularity: 12% [?]

Popularity: 12% [?]

After our recent hiatus (every one has to have time off :-)) it was entertaining to return to one of BadScience.net’s classic subjects - Media Mangling of Stats.

This weeks article (at http://www.badscience.net/?p=230) is about the way news papers fight for headlines by really overdoing the actual data. The headline claims of the number of children using cocaine has “doubled” is based on an increase from 1.4% to 1.9%. Even my basic understanding of maths doesnt see that as a “doubling.”

Well worth a read.

Popularity: 14% [?]

Popularity: 14% [?]