![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
I went to see The Mountain Goats at Koko last night. In its current incarnation its a trio, and it has a much "rockier" feel than do even the recent albums.
I say "in its current incarnation", because The Mountain Goats is eally the front title for John Darnielle, one of the best singer-songwriters around. And yet he was only playing three dates in the UK.
It was one of those few performances where I wished that I had suffered the crush from being at the front, plus the undeniable fact that the front of the stage attracts a significantly higher proportion of twats. Darnielle was on top form, perhaps made more poignant that one never has any idea if or when there will be a chance to see him again.
++++++++++
Unfortunately it became clear yesterday afternoon that I had a cold coming on, and it came into full flow this morning. I felt dreadful, and was extremely glad that I wrote up a lot of stuff yesterday (good stuff, too!) so that my input this morning consisted of only a few last-minute developments.
Then it was back to bed, whence I have just emerged, and whereto I think I shall soon return.
That rest gave me the chance to listen to "More Or Less", the numbers programme on Radio Four presented by FT's undercover economist Tim Harford. These programmes usually have at least one piece that, like Ben Goldacre's "Bad Science", show how numbers are misused or, perhaps more likely, simply misunderstood.
A Guardian piece on Saturday contained these two snippets of information. These, remember, are presented as "facts".
(1) 11m people in the UK are disabled.
(2) "a large majority of those surveyed said they had never had a disabled person to their home for a social occasion."
http://www.guardian.co.uk/commentisfree/2010/sep/06/disability-we-can-all-do-better
Now, applying maths to these two statements is not simple, even though common sense indicates that they don't really add up.
Let's suppose that on average we have three "social occasions" a year, to which we invite an average of 10 people (the majority would be less than this, but some bigger parties would boost it). So let's make that an average of 20 people a year.
Let's suppose that we "know" 150 people (this is a commonly used number). That would mean that we should "know" an average of 25 to 30 "disabled" people. If we picked the 20 from the 150 that we know at random, the chance that none of them would be disabled would be about six per cent. But Naomi Jacobs says that 91% of people never have anyone disable around. That is obviously statistically significant, implying massive prejudice against the disable.
But, hold. It turns out that, while Naomi Jacon's first use of the word "disabled" covers the Department of Work & Pensions research (where people are asked if they suffer from a disability that prevents them living an "ordinary" life), her second use of the word is from research by Scope, where people are asked if they have ever had a "disabled" person in their home for a social occasion.
In other words, the word "disabled" changes its meaning from one part of the article to the next. Now, I'm happy for you to define your terms as you wish, but once the terms are defined, you have to stick to them, else your argument becomes fallacious. Ms Jacobs could as easily have said "There are 11m blue-eyed people in the country, but nine out of 10 of us have never had a blind person into our home", but then it becomes obvious that the first point has no relevance to the second point.
However, Ms Jacobs isn't a logician, statistician or mathematician. And neither are most of her readers. As I presume all of you have spotted, this is a simple mistake of logic, used to provide spurious back-up for a point that needed no back-up.
And it's that which annoys me most, I think, because the standard response to pointing out this error would be, from the Guardianistas, that "it's still a disgrace that so few of us have invited a disabled person into our homes". Indeed, but the revelation that an argument is logically flawed is not the same as saying that the argument is wrong.
This happened recently when someone (he claimed later that it was with foreknowledge) posted on Facebook a memo that purported to show a strategy to short-order one in three McDonald's drive-through customers at a branch in Australia.
It was a hoax (the signature "Robert Trugabe", a facsimile of Robert Mugabe's, should have been a clue) and I pointed this out within a few minutes. But not before another keen dupee (I wish I could have these people on a Scam mailing list) had written something like "I knew it!!!"
She responded to my post that, it didn't matter that the letter was a hoax, because short-ordering did happen.
Is there any point in replying to these people (I don't bother any more -- they might take up poker one day and I hardly want them to get brighter). But, of course, by definition it does happen, because these types of hoaxes only go viral on the Internet because they tap into a chord, they reinforce the person's previous vision of reality. Like, Duh, Obviously, indeed.
So, whenever someone points out a logical flaw in an argument, do not respond "it doesn't matter, because the conclusion is true", beware. You are committing the same error as the two people above. Once an argument has been shown to have a logical flaw, the argument fails. Go away, come back with a rephrased argument, but don't, don't, say "it doesn't matter".
+++++++
In other news, my cold is still dreadful. Could be a no-poker day.
PJ
I say "in its current incarnation", because The Mountain Goats is eally the front title for John Darnielle, one of the best singer-songwriters around. And yet he was only playing three dates in the UK.
It was one of those few performances where I wished that I had suffered the crush from being at the front, plus the undeniable fact that the front of the stage attracts a significantly higher proportion of twats. Darnielle was on top form, perhaps made more poignant that one never has any idea if or when there will be a chance to see him again.
++++++++++
Unfortunately it became clear yesterday afternoon that I had a cold coming on, and it came into full flow this morning. I felt dreadful, and was extremely glad that I wrote up a lot of stuff yesterday (good stuff, too!) so that my input this morning consisted of only a few last-minute developments.
Then it was back to bed, whence I have just emerged, and whereto I think I shall soon return.
That rest gave me the chance to listen to "More Or Less", the numbers programme on Radio Four presented by FT's undercover economist Tim Harford. These programmes usually have at least one piece that, like Ben Goldacre's "Bad Science", show how numbers are misused or, perhaps more likely, simply misunderstood.
A Guardian piece on Saturday contained these two snippets of information. These, remember, are presented as "facts".
(1) 11m people in the UK are disabled.
(2) "a large majority of those surveyed said they had never had a disabled person to their home for a social occasion."
http://www.guardian.co.uk/commentisfree/2010/sep/06/disability-we-can-all-do-better
Now, applying maths to these two statements is not simple, even though common sense indicates that they don't really add up.
Let's suppose that on average we have three "social occasions" a year, to which we invite an average of 10 people (the majority would be less than this, but some bigger parties would boost it). So let's make that an average of 20 people a year.
Let's suppose that we "know" 150 people (this is a commonly used number). That would mean that we should "know" an average of 25 to 30 "disabled" people. If we picked the 20 from the 150 that we know at random, the chance that none of them would be disabled would be about six per cent. But Naomi Jacobs says that 91% of people never have anyone disable around. That is obviously statistically significant, implying massive prejudice against the disable.
But, hold. It turns out that, while Naomi Jacon's first use of the word "disabled" covers the Department of Work & Pensions research (where people are asked if they suffer from a disability that prevents them living an "ordinary" life), her second use of the word is from research by Scope, where people are asked if they have ever had a "disabled" person in their home for a social occasion.
In other words, the word "disabled" changes its meaning from one part of the article to the next. Now, I'm happy for you to define your terms as you wish, but once the terms are defined, you have to stick to them, else your argument becomes fallacious. Ms Jacobs could as easily have said "There are 11m blue-eyed people in the country, but nine out of 10 of us have never had a blind person into our home", but then it becomes obvious that the first point has no relevance to the second point.
However, Ms Jacobs isn't a logician, statistician or mathematician. And neither are most of her readers. As I presume all of you have spotted, this is a simple mistake of logic, used to provide spurious back-up for a point that needed no back-up.
And it's that which annoys me most, I think, because the standard response to pointing out this error would be, from the Guardianistas, that "it's still a disgrace that so few of us have invited a disabled person into our homes". Indeed, but the revelation that an argument is logically flawed is not the same as saying that the argument is wrong.
This happened recently when someone (he claimed later that it was with foreknowledge) posted on Facebook a memo that purported to show a strategy to short-order one in three McDonald's drive-through customers at a branch in Australia.
It was a hoax (the signature "Robert Trugabe", a facsimile of Robert Mugabe's, should have been a clue) and I pointed this out within a few minutes. But not before another keen dupee (I wish I could have these people on a Scam mailing list) had written something like "I knew it!!!"
She responded to my post that, it didn't matter that the letter was a hoax, because short-ordering did happen.
Is there any point in replying to these people (I don't bother any more -- they might take up poker one day and I hardly want them to get brighter). But, of course, by definition it does happen, because these types of hoaxes only go viral on the Internet because they tap into a chord, they reinforce the person's previous vision of reality. Like, Duh, Obviously, indeed.
So, whenever someone points out a logical flaw in an argument, do not respond "it doesn't matter, because the conclusion is true", beware. You are committing the same error as the two people above. Once an argument has been shown to have a logical flaw, the argument fails. Go away, come back with a rephrased argument, but don't, don't, say "it doesn't matter".
+++++++
In other news, my cold is still dreadful. Could be a no-poker day.
PJ
