Biases against science

Professor Mark Pepys says “grossly inadequate” education has left most people “tone deaf” to science.
I fear he’s right. But the problem isn’t confined to schools. There’s a vast number of biases that stop people thinking scientifically.
First, a matter of definition. Science is not merely, or even mainly, a body of facts. If it were, the problem of scientific ignorance would be easily solved.
Instead, the importance of science lies in its method - the way in which theories (stress the plural) are challenged against as much evidence as possible.
In this sense, even scientists often fail to be scientific, as Richard Dawkins and James Watson have recently shown. And many doctors have a notoriously vague grasp of probability.
I reckon there are at least four biases against the scientific method:
1. The power of authority. From infanthood onwards, we’re brought up to believe authority. It’s often sensible to do so. Parents and teachers know more than us. And it’s just impractical to work everything out for ourselves.  But the scientific method requires that we believe not people but the evidence - and, indeed, are sceptical even of that.
In this sense, Lord Rees - president of the Royal Society - was encouraging anti-science when he spoke recently of the “scientific consensus.” You don’t reach the truth through opinion polls.
2. The power of anecdote. People believe single, salient stories more than thousands of statistical data points. Take the question: does the MMR vaccine cause autism? The proper way to answer this is to fill in the four boxes (jab/no jab, autism/no autism) to establish correlations, and to ask: what are the possible mechanisms linking the vaccine to autism?
Instead, people preferred the vivid story: “the son of a friend of a friend had the jab, and a few weeks later seemed to have autism.“ Few asked the scientific questions: how representative is this story? What’s the mechanism?
The media perpetuate this bias. Journalists much prefer the human interest story to dry statistical inference. But you don’t necessarily get to the truth through entertainment.
3. The cult of self-expression. Everyone thinks they “have a right to an opinion”, a views fostered by vox pops and phone-in programmes. But opinion doesn’t matter. What matters is evidence and thought. Proper science is democratic in the sense that it considers all evidence, from whomsoever it comes. But it’s not democratic in the sense that it gives weight to the idle opinion of every passer-by.
4. Overconfidence.  It’s very easy for our confidence in our opinion to grow faster than the evidence. This is especially likely if our achievements in one field win us prizes and esteem. This, I suspect, is part of the reason for James Watson’s unfortunate utterances.
The message here is that it’s not just schools to blame for scientific illiteracy. Indeed, the scientific method is profoundly unnatural - that’s why it took mankind millennia to stumble upon it.

Comments

[Matt Munro]

Totally agree. Post modernism and its hell spawn child, political correctness, with its essential beleif that nothing is "true" and all points of view are equally vaid, is incompatible with the empirical method.

[john b]

...as is traditional 'respect for authority', of course.

[Matt Munro]

Yeah yeah, the nazis used science and they were bad people therefore all science is bad - yawn, yawn yawn, pass the dope, dope.

[alter_ego]

How has James Watson failed "to be scientific"?
http://upload.wikimedia.org/wikipedia/en/thumb/8/88/Sketch-4race-transparent.png/400px-Sketch-4race-transparent.png

[Matt Munro]

If he's racist he's automatically unscientific. The fact that racism is (one possible) social construction of science, and nothing to do with the science itself seems to pass these people by.
Who was to blame for the atom bomb, the bomber crew that dropped it, or the science that made it possible ?

[reason]

Matt Munro...
you clearly misunderstood john b's comment.

As for Watson - well this:
"He said he hoped that everyone was equal, but countered that “people who have to deal with black employees find this not true”".

Is clearly idiotic, as it treats all "black" (what does that mean exactly) people as though they are the same (which anybody who has actually had any dealings with them knows is not true).

As for Dawkins, I want more explaination as to what he was doing wrong. My opinion is that what he said was clearly true, and he acknowledged that it was based on second hand information. Does somebody have objective evidence that it is not true?

[emmanuelgoldstein]

Alter ego might like to read this: http://cogprints.org/230/

[reason]

Just to clarify about James Watson.

I hold it for clearly true that the great majority of exceptional sprinters are of West African origin. I hold it for clearly true that the great majority of exceptional distance runners are of East African origin. It doesn't mean that any randomly selected West African is faster than me over 100 meters, or any randomly selected East African would leave me well behind over 10 kilometers. Even if we were to accept that the average intelligence of Africans is lower than that of Europeans (which is lower than that of East Asians?) it tells me nothing about any individual.

[emmanuelgoldstein]

[The fact that racism is (one possible) social construction of science, and nothing to do with the science itself seems to pass these people by.]

Heroic optimism! (I give you Cyril Burt...)

[Rob Spear]

What if he's not racist, but just claiming there is some correlation between skin colour and intelligence? That sounds like a falsifiable claim to me, and therefore a scientific one.

[reason]

But, Rob Spear, that is not what he is claiming.

[Matt Munro]

emmanuelgoldstein: Psychometrics just says that this or that ability *can* be measured, that's it, that's where the science ends. The rest - the decision to measure it and what that measurement *means* is socially constructed, which is why most (granted not all) criticisms of IQ testing are cuturally based

John B - apologies I realise you were violently agreeing with me

[reason]

Besides which Rob Spear, do you know what "intelligence" means? More than the ciruclar definition, that it is "whatever IQ tests measure". Because otherwise, my guess is that that it may be difficult to falsify because of the lack of reliable metrics.

[Matt Munro]

Reason - it is uncontested that IQ correlates positively with educational outcomes, life chances and incomes. The issue is which is cause and which is effect, and whether intelligence is inherited or "learned". It is possible to train people to improve their IQ scores but that improvement is not reflected in long term academic acheivment, suggesting heritability. On the other hand it's not unknown for people with "average" parents to pass the 11+, (I'm one of them), suggesting a role for nurture. It's most likely that IQ is a function of gene/ environment transactions and you could therefore argue that what IQ realiably measures and predicts is "economic success is a western society".
There are valid criticism of IQ testing, but the idea that it doesn't measure anything usefull isn't one of them

[The SageKing]

Lots to agree with here.

I have to say the first thing to point out is that economics is not a science, if it was economists as you regularly point out, would be able to make certain predictions about the future course of events..just like your hero Marx tried and failed to do, making his theory pseudo-science.

IQ is a very accurate science that does give pretty good predictions of the future BUT
IQ is a bit like a car fuel tank, you might have a big one but you might fill it with meths, which wont give you good performance, and do the engine in.

Genetics is a little bit like logic, you get two pieces of logic that seem to fit perfectly, put them together and they behave in a totally different way as you would have predicted. the law of unknown consequences i should imagine.

A article you might wish to dig up from the economist is the best i have read on the subject.

The evolution of intelligence
Jun 2nd 2005
From The Economist print edition or in a nutshell why are Ashkenazi Jews so brainy?

MORE here, http://en.wikipedia.org/wiki/Ashkenazi_Intelligence#Cochran_et_al.

BTW, as you seem to dismiss Dr Watsons work on the reporting of some clumsy statements he has made, does confirmation bias work on you as well in not taking into account the wider scope of his work and others in IQ?

[melior]

I agree with your larger point, but the slam against Dr. Dawkins is pure-grade kooky religious false martyrdom.

AIPAC does play a dominant role in US lobbying both financially and politically all out of proportion to their numbers. Look it up -- it's a verifiable fact.

This has nothing to do with the particular variety of invisible man they worship, or their race, and Dr. Dawkins never says it does.

Crying 'anti-semitism' in this case is not only baseless slander, it completely misses the factual point Dr. Dawkins makes: "When you think about how fantastically successful the Jewish lobby has been, though, in fact, they are less numerous I am told - religious Jews anyway - than atheists and (yet they) more or less monopolize American foreign policy as far as many people can see. So if atheists could achieve a small fraction of that influence, the world would be a better place."

Please try not to enable deluded idiots when making your point in the future; your identification of genuine 'false science' will retain more respect that way.

[Dipper]

Its disappointing to once again see scientists falling into the IQ trap. Intelligence is not like height, and crass attempts to measure and analyse it are always an embarrassment.

[reason]

Matt Munro...
you seem to have misunderstood my comment. I meant to imply that the statement that would be testable would be that there is a correlation between skin pigment and IQ scores not that there is a correlation between race and intelligence. (i.e. Replacing poorly defined and unmeasurable terms with measurable ones.)

[emmanuelgoldstein]

Matt Munro,

[emmanuelgoldstein: Psychometrics just says that this or that ability *can* be measured, that's it, that's where the science ends. The rest - the decision to measure it and what that measurement *means* is socially constructed, which is why most (granted not all) criticisms of IQ testing are cuturally based]

Don't quite know what you mean.

Intelligence is a vague, if familiar, concept of great interest.

There's a long chain of claims here: (i) that intelligence is measurable, (ii) that IQ is a good measure of intelligence, (iii) that intelligence is highly heritable, and (iv) that blacks (loosely defined) have lower intelligence than everyone else. (Additionally, (v) that this intelligence gap is probably insurmountable; at least, that's one widely-drawn conclusion (Rushton, Jensen, and now, apparently, Watson) from the surmise that there are genetic determining causes for measured IQ, and hence racial differences in measured IQ).

Anyway, (at least some) psychometricians (again, Rushton et al.) seem to see no difficulty whatever in arguing that (i), (ii), (iii), and (iv) can, essentially, be read directly off of the data. That does not appear to be 'social construction' in any relevant sense, so it's unclear to me why you'd think social construction is the issue here.

[Rob Spear]

So, to summarize, Watson is unscientific and a racist because he used the word "intelligence" instead of "IQ" in a newspaper interview?

[Kalmz]

reason wrote...

As for Watson - well this:
"He said he hoped that everyone was equal, but countered that “people who have to deal with black employees find this not true”".

Is clearly idiotic, as it treats all "black" (what does that mean exactly) people as though they are the same (which anybody who has actually had any dealings with them knows is not true).
----

What's clearly idiotic is your misreading of Watson. Where do your "all" and "same" come from? He's talking about averages: some whites are LESS intelligent than some blacks, but on average whites are more intelligent than blacks -- and less intelligent than East Asians and Ashkenazi Jews.

[reason]

Kalmz...
sure, my amazing instanteous statistical analysis machine (brain) automatically deduces averages on the basis of personal contact. The statement IS clearly idiotic - he could have said exactly the same (for instance) about people from red states (some of them are inferior to those from blue states). It is meaningly trying to draw inferences about entire groups from experience with single individuals.

[Term papers]

Totally agree with the above comment ..!