Science is not what it used to be

[petrossa]

Low intelligence among top heart health risks

Wed Feb 10, 2010 6:50am GMT

By Kate Kelland

LONDON (Reuters) - Intelligence comes second only to smoking as a predictor of heart disease, scientists said on Wednesday, suggesting public health campaigns may need to be designed for people with lower IQs if they are to work.

http://uk.reuters.com/article/idUKTRE61900L20100210

Sofar it's just a journalist making it up as she goes along, since the article was published on july 15, 2009. But, well it's a great headline and probably she had a deadline and nothing to write.

But the study itself, that's so absurdly bad my toes curl.

IQ explains some of the difference in heart disease between people of high and low socio-economic status

MEDICAL RESEARCH COUNCIL MEDIA RELEASE

MRC/43/09

Embargoed: 00.01 hrs London time (BST) Wednesday 15 July 2009

A unique study looking at the difference in cardiovascular disease (heart disease and stroke) and life expectancy between people of high and low socio-economic status has found that a person?s IQ may have a role to play.

http://www.mrc.ac.uk/Newspublications/News/MRC006209

In short the paper finds a correlation between heartfailure and being stupid.

They found that IQ explained more than 20% of the difference in mortality between people from socio-economically disadvantaged backgrounds compared to those from more advantaged backgrounds. Importantly, this was in addition to the classical, known risk factors for heart disease, such as smoking and obesity.

How did they come to this conclusion?

Dr David Batty, who led the research, said: ?We already know that socio-economically disadvantaged people have worse health and tend to die earlier from conditions such as heart disease, cancer and accidents. Environmental exposures and health-related behaviours, such as smoking, diet and physical activity, can explain some of this difference, but not all of it. This raises the possibility that, as yet, unmeasured psychological factors need to be considered. One of these is intelligence or cognitive function, commonly referred to as IQ. This measures a person?s ability to reason and problem solve. IQ is strongly related to socio-economic status.?

That's science folks. A bland assumption correlating unproven to be related circumstances and hey presto a nice headline for a newspaper a year later and a paper published.

Did Dr. Batty for example ever consider that one can be very intelligent but incapable the operate socially in an acceptable way. Or that quite often intelligent children on standard schools are badly performing students. That people with socio-economic problems are frequently people with mental problems but have normal range IQ's. That IQ is only strongly related to?socio-economic status in the sense that there are no lower IQ people who beget work requiring a high IQ, but not in the other sense. Many quite normal or above normal people have a low socio-economic status.

Science like this permeates all disciplines. Imho it's because scientist get higher payscales mostly based on their publication rate. This pushes for quantity rather than quality.

[lamminium]

You... read the whole article, right?

...?(i) intelligence might lead to greater knowledge about how to pursue healthy behaviours; (ii) intelligence may ?cause? socioeconomic position, i.e. more intelligence leads to more education, income, occupational prestige . . .; and (iii) intelligence may be a marker for something else, and it is that something else, early life exposures, for example, that leads to mortality.? They speculate as to whether conditions in early life, ranging from foetal programming to parental interest in a child?s education might influence both IQ and subsequent risk of disease, and they suggest that a combination of the second and third explanations ?have much to recommend them and point to interventions at societal level?.

[...]

Commenting on the public health implications of his study, Dr Batty and his colleagues write: ?Our findings suggest that measured IQ does not completely account for observed inequalities in health, but, probably through a variety of mechanisms, may quite strongly contribute to them. This implies that efforts to reduce inequalities should continue to be broadly based, including educational opportunities and interventions directed at early life. . . . It may be that individual cognition levels should be considered more carefully when preparing health promotion campaigns and in the health professional?client interaction.?

[iamawesomewicked]

You're right petrossa.. Science isn't what it used to be.. it's MUCH, MUCH, MUCH better.

[zhangm Supervisor]

There is a significant correlation between IQ and expected lifespan. The term correlation has a very specific meaning in scientific literature, and it is a much weaker assertion than the stupid general public normally takes it to mean.

If the best you can do is to interpret science through the mouthpiece of news media, and also misinterpret the terminology, perhaps you should stay away from coming to unfounded conclusions yourself.

[carmatic]

There is a significant correlation between IQ and expected lifespan. The term correlation has a very specific meaning in scientific literature, and it is a much weaker assertion than the stupid general public normally takes it to mean.

If the best you can do is to interpret science through the mouthpiece of news media, and also misinterpret the terminology, perhaps you should stay away from coming to unfounded conclusions yourself.

the way i see it, a correlation is more like a clue to the answer, rather than the answer itself...

[petrossa]

You... read the whole article, right?

Silly question, you should know better by now, ofcourse i did. I read the original paper in 2009 already. Dr. Batty is a very good name for him.

What counts for science nowadays would get laughed out of college in the 60's.?

Our findings suggest that measured IQ does not completely account for observed inequalities in health, but, probably through a variety of mechanisms, may quite strongly contribute to them.

So he says in his paper itself, if you'd read it you'd have noticed, that IQ is very important and a definitive factor, and as conclusion says exactly the same but now larded with if's, but's, maybe's and perhapses.

Science. Sigh.

[lamminium]

Silly question, you should know better by now, ofcourse i did. I read the original paper in 2009 already. Dr. Batty is a very good name for him.

What counts for science nowadays would get laughed out of college in the 60's.

Our findings suggest that measured IQ does not completely account for observed inequalities in health, but, probably through a variety of mechanisms, may quite strongly contribute to them.

So he says in his paper itself, if you'd read it you'd have noticed, that IQ is very important and a definitive factor, and as conclusion says exactly the same but now larded with if's, but's, maybe's and perhapses.

Science. Sigh.

Would you continue/have a reason to continue an investigation if you just brush off the correlation as a mere accident? Certainly not.

As with any study we've discussed, one study doesn't set the universal trend but serves to alert the community to an interesting case that is worth looking at.

The author did exactly that. If later the community finds it implausible, at least we didn't miss something that could be important.

If you're not happy with science, I won't try to convince you again. Not everyone can be pleased and I've learnt to accept that.

[petrossa]

Would you continue/have a reason to continue an investigation if you just brush off the correlation as a mere accident? Certainly not.

As with any study we've discussed, one study doesn't set the universal trend but serves to alert the community to an interesting case that is worth looking at.

The author did exactly that. If later the community finds it implausible, at least we didn't miss something that could be important.

If you're not happy with science, I won't try to convince you again. Not everyone can be pleased and I've learnt to accept that.

Why would you want to convince me of anything?. I just had it with semi-scientific studies running around with correlations. As i have conclusively proven in my fat is not bad piece correlations mean exactly nothing.

Astrology is also correlation.

[lamminium]

And with time, we find out that astrology is not reliable. That goes back to what I said earlier...

With fat, you didn't prove anything "conclusively" but merely presented a paradox which didn't really do more than tell us that the fat may produce some beneficial effect (but considering the larger implications that obesity brings, one can hardly agree that excessive fat [note the word "excessive"] is good).

[Farstrider]

Why would you want to convince me of anything?. I just had it with semi-scientific studies running around with correlations. As i have conclusively proven in my fat is not bad piece correlations mean exactly nothing.

Astrology is also correlation.

You do not really seem au fait with the definition of the word correlation! Here we go!

1. mutual relation of two or more things, parts, etc.

2. the act of correlating or state of being correlated.

3. Statistics. the degree to which two or more attributes or measurements on the same group of elements show a tendency to vary together.

4. Physiology. the interdependence or reciprocal relations of organs or functions.

5. Geology. the demonstrable equivalence, in age or lithology, of two or more stratigraphic units, as formations or members of such.

[petrossa]

You do not really seem au fait with the definition of the word correlation! Here we go!

1. mutual relation of two or more things, parts, etc.

2. the act of correlating or state of being correlated.

3. Statistics. the degree to which two or more attributes or measurements on the same group of elements show a tendency to vary together.

4. Physiology. the interdependence or reciprocal relations of organs or functions.

5. Geology. the demonstrable equivalence, in age or lithology, of two or more stratigraphic units, as formations or members of such.

Stars aligning at a certain time of birth = correlation. I guess you don't fully understand what astrology does.

[Kirkburn]

Astrology is also correlation.

Stars aligning at a certain time of birth = correlation. I guess you don't fully understand what astrology does.

In what strange way is that correlation in any kind of scientific sense? It sounds like saying I'm correlated with my laptop at the moment.

The "correlation" that it should be pretty obvious is being talked about is statistical. Stars aligning (with what?) at birth is not.

Correlation when applied to astrology would be related to any correlation of predictions compared to real life. (Of which there is none, of course)

[zhangm Supervisor]

Actually, Jupiter's orbiting the sun is correlated with births, with an N in the billions by now. On the face of it, this correlation seems to suggest some relationship between births, and Jupiter's orbiting the sun.

When you start examining it with anything beyond a superficial mind however, you start to see that there may not be any special relationship between Jupiter's orbiting the sun and people being born. One problem with the original assumption is that while Jupiter does orbit the sun when people are being born, other things also orbit the sun at the same time. Jupiter's orbiting the sun is not something unique that occurs when people are being born, and, in fact, upon second observation one realizes that Jupiter's orbiting the sun is also correlated with people not being born.

So how is it possible that correlation can ever be informative? Well, correlations can be used to create model systems, upon which testable hypotheses can be grafted. With the example above, one can hypothesize from the original correlation that Jupiter's orbiting the sun is required for people to be born. A simple way to test the hypothesis is to remove Jupiter, and observe whether births continue. With regard to the issue of IQ, we can go to schools located in well-off areas or impoverished areas, and attempt to locate children with high and low IQs. We can then monitor them over several years, collecting information about jobs, socioeconomic status, health, and lifespan. The data that we gather suggests that while socioeconomic conditions during development do play a significant role, IQ is also a factor. People with higher IQs are more likely to take advantage of opportunities to improve their own status. Subsequently, they become wealthier, and have better access to nutrition, health care, recreation, and stress relief.

Having a higher IQ doesn't cause good cardiovascular health, resistance to arterial disease, etc, just like it doesn't cause people to be wealthy. Certainly the research doesn't state this, as it would be absurd. The research simply states that higher IQ is correlated to better health and offers a (quite reasonable) model for why this may be the case. I honestly don't get the OP's gripe about this, and I suppose my understanding may be impaired by his poor ability to communicate in coherent English. The researchers found a statistically significant correlation between IQ and health risks, and offers a plausible model for why such a correlation exists. Someone will follow up on it, collect more data, and either support the hypothesis or refute it.

[petrossa]

Having a higher IQ doesn't cause good cardiovascular health, resistance to arterial disease, etc, just like it doesn't cause people to be wealthy. Certainly the research doesn't state this, as it would be absurd. The research simply states that higher IQ is correlated to better health and offers a (quite reasonable) model for why this may be the case. I honestly don't get the OP's gripe about this, and I suppose my understanding may be impaired by his poor ability to communicate in coherent English. The researchers found a statistically significant correlation between IQ and health risks, and offers a plausible model for why such a correlation exists. Someone will follow up on it, collect more data, and either support the hypothesis or refute it.

Voice of reason, finally. Thanks you're a rare species.

Although this makes it an ad hominem: ?I suppose my understanding may be impaired by his poor ability to communicate in coherent English.?I could just as well state that your understanding is impaired by your poor ability to understand English.

Leave those phrases out, they may satisfy your ego, but don't add to the overall impression one gets of your capability to be objective.

I'll better explain my gripe with this faulty model.

First of all there's no mention whatsoever what's the cutoff point in the whole paper. I.e. what constitutes a high/low IQ. Which immediately comes to the core of the whole bizarre model: IQ has no meaning. IQ tests IQ and as such has no bearing on the persons capabilities in general, but in a very limited feild, namely: someone with a high IQ has higher testresults than someone with a lower IQ.

All IQ tests prove is how well one does an IQ test.

Anyone who has ever participated in daily social/work life should have noticed that mediocre IQ people having 1 talent (for example leadership) can score high on the socio-economic ladder, whilst borderline lunatics with IQ's off the scale can live in squalor without any social interaction whatsoever.

The 'model' DR Batty (what's in a name) concocted reeks of a paternalistic view based on his own perceived social standing and limited social interactions. And as such completely without merit.

So, lower 'classes' who are the 'lower' classes by reason of insufficient mental capabilities are more likely to live unhealthy lives because they lack the funds to properly nourish themselves. So they eat not unhealthy because they are too stupid to eat healthy, but because they've no cash.

So the perceived correlation between IQ is false, the real correlation is not enough money. And the lack of money is the CAUSALITY which augments the healthrisks.

Smart people gorging on luxurious meals and fine wines are just as likely to suffer health effects, so there's also a correlation between too much money/IQ and healthrisks.

All in all the paper is a perfect model for determining the bias of the researcher(s).

[Kirkburn]

<snip>

You seem to be putting way too much of your personal opinions into your conclusions.

e.g. "Smart people gorging on luxurious meals and fine wines are just as likely to suffer health effects, so there's also a correlation between too much money/IQ and healthrisks." Did you do a study on this? Isn't this exactly the kind of thing you'd complain about - people making conclusions without justification? "Common sense" =/= valid scientific data.

You also complain about ad hominems, and then perform one against the scientists name ("DR Batty (what's in a name)"). Jeez!

I feel like you should perhaps reread your posts after/during posting because you do present them in a somewhat difficult way to understand.

You make some good points about the applicability of IQ, but it's not as if IQ has no relation (or indeed, correlation) to other factors. There are of course outliers, but exceptions don't prove a general correlation wrong. High IQ is correlated with high intelligence (which is generally correlated with success), and vice versa.

[petrossa]

[quote name=Kirkburn' date='16 February 2010 - 16:22

' timestamp='1266333776' post='592245578]

You seem to be putting way too much of your personal opinions into your conclusions.

e.g. "Smart people gorging on luxurious meals and fine wines are just as likely to suffer health effects, so there's also a correlation between too much money/IQ and healthrisks."

I'll get back on that. I've in my extensive research material somewhere studies linking wealth and CVD inteventions

Did you do a study on this? Isn't this exactly the kind of thing you'd complain about - people making conclusions without justification? "Common sense" =/= valid scientific data.

You also complain about ad hominems, and then perform one against the scientists name ("DR Batty (what's in a name)"). Jeez!

An ad hominem is not discussing the context but the person, a wordplay on a name is a jest. I fully discussed the content and found it batty. Hence the quip, which as as far from an ad hominem as possible.

I feel like you should perhaps reread your posts after/during posting because you do present them in a somewhat difficult way to understand.

My posts are in too formal english because i suffer from Aspergers syndrome. I can't write any other way. Get used to it or ignore it. Don't ridicule it.

You make some good points about the applicability of IQ, but it's not as if IQ has no relation (or indeed, correlation) to other factors. There are of course outliers, but exceptions don't prove a general correlation wrong. High IQ is correlated with high intelligence (which is generally correlated with success), and vice versa.

Weirdly enough i consider myself capable to have my own opinion. And as such all that get's put in rebuttal to my personal insights (based on years of absorbing and even understanding knowledge) is some hautain out of hand dismissal because 'a scientist' (who'm you obviously never read anything? about going by your responses) puts forward a very wobbly model.

The model is so full of holes i wouldn't know where to begin. So i took the most obvious one: IQ doesn't mean much except that it tests the IQ. It's an amusing method for selfmasturbation, you can have whole clubs with them.

But in the end premortality numbers are not significantly lower for people which scored good on any test be it IQ test or their drivers test.

In business good managers need no high IQ scores, they need a strong personality and the quality to make fast decisions.

That alone makes IQ as model for health completely absurd.

[zhangm Supervisor]

Although this makes it an ad hominem: I suppose my understanding may be impaired by his poor ability to communicate in coherent English. I could just as well state that your understanding is impaired by your poor ability to understand English.

No, I was serious. I went back to your first post and tried to better understand what you were saying. I found this:

A bland assumption correlating unproven to be related circumstances and hey presto a nice headline for a newspaper a year later and a paper published.

It doesn't make sense, so I stand by my statement.

So the perceived correlation between IQ is false, the real correlation is not enough money. And the lack of money is the CAUSALITY which augments the healthrisks.

Again, you display a fundamentally flawed conception of what a correlation is. The correlation is not false - it is real, it exists, they observed it. You are still somehow confusing correlation with causation, and for the last several posts, we've been trying to point this out. Also, money is not necessarily a causative factor in poor health. One simple example is crack cocaine, heroine, or tobacco. All are expensive habits, i.e., you dump large amounts of money into maintaining them. In the end however, they do not cause you to have good health, and perhaps this might surprise you, but they cause the opposite.

The researcher's model simply proposes that people with higher IQs have more intellectual capital that they can use to, say, get more money, or move themselves out of the slum in which they grew up. They're more likely to do well in life, make better choices, not believe in homeopathy, etc. It doesn't have much to do with money, as you pointed out.

Smart people gorging on luxurious meals and fine wines are just as likely to suffer health effects, so there's also a correlation between too much money/IQ and healthrisks.

You're going to define smart people as those who make dumb decisions in order to prove your point? If you start with a false assumption, you can prove anything, but a faulty proof like the one above is worthless. The whole trend associated with people that have higher IQs is that they make better life choices - eating healthy, sleeping early, avoiding hard drugs, getting healthcare. It is not gorging themselves to the point of detriment. :rolleyes:

And I highlight the term trend, because the reason that this paper's observations have value is that they apply in the majority of cases. There are obviously exceptions, but harping on the exceptions doesn't do anything against their argument that this is a general observable trend.

All in all the paper is a perfect model for determining the bias of the researcher(s).

It seems like this thread is a perfect opportunity for you to learn about your conceptual shortcomings and correct them, but you aren't doing it.

[Farstrider]

I really hope that Hans Asperger knew how to correlate! :laugh:

I have to admit that I do find it rather amusing how serious and uptight some people get about subjects that ultimately mean diddly squat in the long run! ;)

[petrossa]

Again, you display a fundamentally flawed conception of what a correlation is.

Read this and tell me again i don't know what correlation is:?

The disease FAT does not exist

19 August 2009

52 Comments

Posted by +petrossa

Correlation and Causality ran wild

We are overrun with studies, not a day goes by or we learn another result of a study. Alcohol is good for the heart, alcohol causes cancer. Smoking is unhealthy, smoke is good against dementia. The climate warms, the climate cools, the climate does both at the same time.

Why one wonders, are so many studies contradictory?

The underlying reason for this type confusing headlines is that many scientists confuse correlation with causality. They do a study, find a link between the one and the other (correlation) and then they just assume causality.

A point in case:

Conclusion of a study: Meat causes cancer, especially red meat.

Correlation: more meat more cancer.

Causality: Almost all meat, but certainly red meat, is roasted/baked/grilled. The process of roasting/baking/grilling produces aromatic esters which are known long time to be carcinogenic. Red meat is nearly always roasted/baked/grilled therefore one sees more cancer in people who eat red meat.

The conclusion of the study would then have to be: Fried food increased the risk of cancer.

I will try to make the dangers of such correlation/causality clear by means of an example how this process takes place.

The rest: https://www.neowin.net/forum/blog/316/entry-3088-the-disease-fat-does-not-exist/

[petrossa]

It seems like this thread is a perfect opportunity for you to learn about your conceptual shortcomings and correct them, but you aren't doing it.

It seems more like a way for you to project your perceived superiority be demeaning that of another. Sorry, but i don't accept you as my superior, nor your arguments having any qualitative content making them so.

Smugness is not an argument.?

Refuting what i've said is. The more adjectives you need to demean the others position the less likely it is to be of any consequence. Your writings are just one long ad hominem inter-spaced with adjectives adding NULL information.

A correlation is anything that you want it to be, thats why it means co relation. There's a correlation between shoesize and beer consumption. Doesn't mean anything at all but it's there.

Dr Batty proposes a correlation between IQ , a unit without meaningful content, and socio economic status, a wholly subjective state of life.

Relating the two can only lead to nonsense, or nonscience if you want as neither are objectively quantifiable.

It' beyond comparing apples and pears, it's more like comparing the concept of food to the meaning of life and finding a correlation between apple eaters and people who believe life has any sense.

[Kirkburn]

The conclusion of the study would then have to be: Fried food increased the risk of cancer.

I will try to make the dangers of such correlation/causality clear by means of an example how this process takes place.

I think we have an issue here of how the media represents reports and what the reports are really trying to say. It's fair to say some researchers play into the media's hands since it gets them exposure, but few self-respecting journals would ever publish a paper that mixes up causality and correlations.

Scientific papers will generally say that they found a correlation between consumption of fried food and relative cancer risk. The media, however, will interpret this as "fried food causes cancer" because it makes for a more interesting story. Some researchers will go along with this, foolishly thinking its helping. (Hell, the correlation could be suggesting that cancer causes fried food consumption - unlikely, of course)

Again, you say IQ has no meaningful content, but that's not strictly true. I doubt there is no correlation between IQ and general "intelligence". Why is "socio economic status" entirely subjective?

There's a correlation between shoesize and beer consumption.

You propose this as an absurd correlation, but it doesn't have to be. Larger shoe size = adult, men - which is indeed those likely to drink more. If you had a study that found a correlation between the two (which is possible, but unlikely to be strong), you can still make inferences - i.e. that larger shoe size indicates something which affects beer drinking.

Besides, there are varying strengths of correlation - it's not a case of exists or not exists.

[petrossa]

I think we have an issue here of how the media represents reports and what the reports are really trying to say. It's fair to say some researchers play into the media's hands since it gets them exposure, but few self-respecting journals would ever publish a paper that mixes up causality and correlations.

Scientific papers will generally say that they found a correlation between consumption of fried food and relative cancer risk. The media, however, will interpret this as "fried food causes cancer" because it makes for a more interesting story. Some researchers will go along with this, foolishly thinking its helping. (Hell, the correlation could be suggesting that cancer causes fried food consumption - unlikely, of course)

Again, you say IQ has no meaningful content, but that's not strictly true. I doubt there is no correlation between IQ and general "intelligence". Why is "socio economic status" entirely subjective?

Well that's how i wrote the OP actually. I started of by mocking the Newspaper headlines. After that is started to undermine some of the good Dr Batty's preconceptions. He brings up socio-economic status. That's a completely individual thing.?

If he had said: the people with more money get better care, it would mean something because from personal experience i can avow this to be true.

But people of average intelligence can and do become quite rich, You just need to have a talent in demand and you're home free. Whereas people with intelligence off the scale rarely get superwealthy. Nobel gained a fortune, Einstein didn't.

[Kirkburn]

Well that's how i wrote the OP actually. I started of by mocking the Newspaper headlines. After that is started to undermine some of the good Dr Batty's preconceptions. He brings up socio-economic status. That's a completely individual thing.

If he had said: the people with more money get better care, it would mean something because from personal experience i can avow this to be true.

But people of average intelligence can and do become quite rich, You just need to have a talent in demand and you're home free. Whereas people with intelligence off the scale rarely get superwealthy. Nobel gained a fortune, Einstein didn't.

You're extrapolating from personal experience or knowledge again, which is a dangerous road to take. Correlation deals with trends and the majority, not outliers and exceptions.

I don't quite understand why "socio-economic status" is necessarily a subjective term? I would have thought most implementations of measuring it intentionally avoid relying on subjective factors. Whether you find it important is subjective, but not really the measurement itself. e.g. wealth is an indicator of it, even if you don't personally care too much for money.

[petrossa]

a socio-economic status is , as the word status implies a relative concept. It conveyes 'status as compared to others' as such it's not a quantifiable emperically determined entity. Using this and an totally disqualified test 'result' as core for a proposed model isn't science but arrogant assumption.

[Kirkburn]

I think I should just link http://en.wikipedia.org/wiki/Socioeconomic_status at this point. It is quantifiable, because the factors that make it up are quantifiable.