to think that people who do not understand that an anecdote is not the same as a statistic are stupid?

[turningvioletviolet]

Listening to Jeremy Vine (for my sins). Topic is 'breastfed children are more clever than those who aren't' (which is neither here nor there), based on a Brazilian study.

But the amount of people who are texting in saying 'my child wasn't breastfed and is top of the top school in the country so therefore the study IS NOT TRUE' or conversely that they did breastfeed and their child is thick as pig s**t so the study IS NOT TRUE.

Your anecdote means NOTHING.

And so ends my daily rant at the Jeremy Vine show.

"You can prove anything with facts" - do you mean 'you can prove anything with statistics'? Well, no, you can't. Well-founded scepticism about badly-designed studies then further misinterpreted by journalists is a good thing. Total rejection of the conclusions of all science and mathematics by saying that they can essentially make up what they like is, well, just fingers-in-ears ignorance, sorry.

"You can prove anything with facts"

Yes, that is the idea, with facts!

Yabu, I think. There's no point trying to make this a science argument when clearly it's a BF argument.

When people look around them and see no jeffing difference between bf/ff babies, it's very wearisome to be told but the bf kids do better.

Really? I'll bring my kid up to be a brainbox like me, and what he ate five years ago will be such an insignificant factor as to make reports like this laughable.

Also they highlight tiny, tiny margins. If the margins were in any way material, why can't we see the evidence all around us? I don't need fresh reports telling me the dangers of smoking or the benefits of taking anti biotics to beat infection. I see daily illustrations of those truths.

Yet I see and hear of no measurable difference between bf and ff kids and adults.

And if we discard anecdote on the grounds that it fails scientifically, we might as well shut mumsnet down and post everybody a peer fucking reviewed text book about parenting.

There's your data, now get on with it and shut up.

Doctors used to advertise smoking as good for your lungs. The evidence was all around people then, but until proper statistical studies were done, nobody noticed it. You might not have noticed a halving in the SIDS rate since the back to sleep campaign, since it's such a rare event, but thank god for the statisticians that did all those maths for us. You can't just see the evidence for small population-level differences in outcomes, but that isn't to say they aren't important.

We need more statistics teaching in school! I'm no statistical genius, but blimey, some of the rubbish on this thread...

What you have to acknowledge is that a proper, scientific, research project is a completely different thing from asking your mates down the pub / coleague in the office/neighbours in your road but poor media reporting will still just say

"Research has shown...."
"A survey out today says....."

which is why so many people will question 'statistics' or 'survey' results, because the vast majority that are quoted in popular media (be that a magazine or a newspaper or local radio or whatever) are actually poorly put together questions, or are 'sponsored' by a company that wants a certain answer, or have all been asked of one particular demograph, or the sample was so small as to be scientifically insignificant. All of which will be challenged by people with a little more than average intelligence or education. A lot of 'statistics' are completely flawed.

Yes, but then you should be criticising that, not ALL statistics!

Yes it annoys the life out of me.

'well my blessed son was...'

it doesn't matter. Your son is just a part of the overall statistic. And the statistics say THIS

Statistics are only meaningful if you know the context in which they were gathered and by whom, and whether there are other factors which might impact on the study.

I'm sure the Today programme on R4 reported that one of the reasons this study was particularly important was because it had gone much further than other studies to remove the impact of non BFing influences on the results.

"Dr Bernardo Lessa Horta, from the Federal University of Pelotas in Brazil, said his study offers a unique insight because in the population he studied, breastfeeding was evenly distributed across social class - not something just practised by the rich and educated.

Most of the babies, irrespective of social class, were breastfed - some for less than a month and others for more than a year.

Those who were breastfed for longer scored higher on measures of intelligence as adults.

They were also more likely to earn a higher wage and to have completed more schooling."

It was also reported that the results were not conclusive.

Found the Guardian article

www.theguardian.com/science/brain-flapping/2015/mar/18/breastfeeding-raises-iq-worrying-questions

As far as I tell tell the solution to all of this is massively more and better teaching about statistics. If someone could actually turn around when they were reading this stuff and say, "well they didn't correct for maternal education so it's flawed" rather than "my neighbour's nephew's chauffeur went to university" we would really be better off.

If you can, as some do, find the original study and at least read the abstract, people would actually know what the study is designed to do, rather than the soundbite and media decides people should hear.

Just understanding the difference between a study looking at correlation and a study with control groups and experimental method would be a start.

It has only ever meant "a bf child has a few extra IQ points than he would have had if he had been ff instead" and IIRC has never been claimed to mean more than 3-5 IQ points anyway. IQ test results for an individual can vary that much, so it isn't a significant increase.

I understand the science, I understand the statistics, and since I bf all my babies into the "extended" period I ought to feel vindicated. But I don't. "Than if he had been bf" is a completely meaningless comparison in the real world.

People don't understand statistics. A 1/14m chance on the Lottery is good enough "I could win" but a 1/100 chance of contraceptive failure "could never happen".

I like it when news reporters translate to real world parallels. In this case it would be helpful to say eg "bf instead of ff could result in a 3-point higher IQ, which is equivalent to the difference between this rocket scientist and this brain surgeon" - equally you could spin it as "...which is equivalent to this doctor who scraped his grades at A-Level, and this burger flipper who only just missed his".

I was pondering about this some more. The other topic that gets regularly trotted out as study vs anecdote is whether mass immigration depresses wages. All of the credible econometric work in the UK says that immigration doesn't have a depressive effect on local wages, but the general population refuses to accept this, in large part because they have experience (or anecdote, including from the media) of work that used to be done by the UK population and is now largely the preserve of often Eastern European workers (farm labouring, for example)

I think this is because the methodology of studies conducted means that they have not looked a specific sectors / parts of the country, and because there are other factors at play that they haven't been able to control for, but it's an interesting and apparently contradictory position

But the study wasn't a comparison of FF vs. BF outcomes.
It was studying outcomes for children over a 30 period (i.e. long enough to be able to establish their careers and earning potential as well as IQ) and seeing if there was a correlation between the length of time they were BF and "success" (as defined by education, income and IQ).

"They found that all the breastfed babies had greater intelligence, as measured by a standard IQ test, had spent more years in education and had higher earnings. But the longer they had been breastfed, the greater the benefits. Children who had been breastfed for 12 months had an IQ that was four points higher than those breastfed for less than a month, had nearly a year’s more schooling and earned around £70 a month more – about a third more than the average income level." (from the Guardian article). Which seems pretty real world to me.

Earning a third more a month is certainly real world and tangible!

This is a large scale study done over 30 years. Not a half assed small thing you can confidently ignore.

I didn't read it, MrsCakes, because I cba but I agree £70/w is a useful real-world explanation. So why wasn't that the headline?

I have no idea - I'm not a journalist and am often bemused as to what they seem to think is the real story.

Sorry - you can prove anything with facts is a Stewart Lee quote... Point he's making is people's tendency to dismiss factual evidence when it doesn't suit what they want to believe...

YANBU.

I get frustrated when people dismiss the results of studies, (especially large ones with thousands or tens of thousands of individuals) showing that people in group A are more likely on average to be cleverer / healthier / whatever - with the reasoning that because they know people who don't fit with that average, the statistics and the study must be wrong.

Completely missing the point that statistics talk about the likelihood of a particular outcome, not absolute certainties.

There's now an entire other thread with someone claiming that finding a non-wealthy person in Brazil who was BF will prove that the study is "cack".

Lest this seem like I am only annoyed at statistics abuse in a pro-BF way, I also get REALLY PISSED OFF at BF mums who claim that their children never get sick "just because they're breastfed!". THAT'S NOT HOW IT WORKS.

Lol leedy stay away from the anti-vaccination threads you head would break through the table by the end!

YANBU this winds me up too but human beings are frequently irrational and the power of stories has a much stronger grip over our psyche than statistics.