Jan and Klodi's Party Bus - part II **off topic discussion**

Chuchote

Fian wrote: »

I get really irritated with the constant "correlation = causation" you see in the media nowadays.

But surely there is an element of causation in studies like these, or the Doll & Hill ones and the Surgeon-General's Report that proved conclusively that smoking caused lung cancer? This is the purpose of such studies?

Weepsie

Booked a bike onto Irish Rail journey last week. Get on train, first bike rack has bags thrown all over it.

They helpful staff pointed me to the other carriage and guaranteed it'd be empty.

I picked a seat beside the bike originally, but they then assigned me a seat the next carriage down which is a bit daft. Someone was in my seat so and there were vacant seats beside the bike.

The website is definitely improved, but a bit of logic should be applied to these types of bookings.

Fian

Chuchote wrote: »

But surely there is an element of causation in studies like these, or the Doll & Hill ones and the Surgeon-General's Report that proved conclusively that smoking caused lung cancer? This is the purpose of such studies?

No, or rather not necessarily.

The hypothetical :

"study finds over 98% of people who always fly first class are wealthy and literate"

is a great example.

There is a correlation - I have no doubt those statistics are accurate. This does not mean flying first class causes people to become wealthy. Nor does it mean flying first class causes people to be educated. what it means is that people who are wealthy tend to be educated. People who only fly first class tend universally to be wealthy. People who are so wealthy as to be able to fly first class all the time tend even more firmly than ordinarily wealthy people to be educated. But there is no element of flying first class causing wealth or education.

The obesity one is perhaps a little less obvious, but I have no doubt that the correlation between cycling to work and reduced obesity has very little to do with cycling causing a reduction in obesity - rather that people who are obese are unlikely to cycle.

The obesity is causing a sample selection distortion rather than the cycling causing a massive reduction in obesity. People who choose to exercise and cycle tend to also watch their diet, People who are obese are embarrassed at being laughed at for not being able to get up a hill or just being ridiculed for being on a bike so are discouraged from cycling, people who are obese are unlikely to be enthusiastic about physical activities, because they are harder and more sweaty for obese people and because obese people tend to have less energy.

Yes cycling is decent exercise, yes exercise does help with weight - to a limited extent. But there is no way in hell that cycling reduces your risk of obesity by 39%, still less that it causes a 39% reduction in obesity.

Instead it is that being obese reduces the likelihood of you being a cyclist by a significant extent.

edit: but of course lots of these studies make good clickbait so get picked up by the media, academics need to pad the number of studies they publish and so it is not too surprising you see enough of these to get irritated by them.

Chuchote

And how would you apply this to the series of studies mentioned, Doll & Hill and the Surgeon General's Report? People who have lung cancer are more likely to take up smoking?

Fian

Chuchote wrote: »

And how would you apply this to the series of studies mentioned, Doll & Hill and the Surgeon General's Report? People who have lung cancer are more likely to take up smoking?

I wouldn't.

Just because some correlations also involve causation does not mean all of them do.

I don't doubt that smoking causes lung cancer, because smoking introduces carcinogenic toxins into your lungs. Therefore it is not surprising that there is an increased incidence of lung cancer among smokers.

If you had a random sample of 10,000 people. If you organised them into buckets according to age, gender, weight, race and income. If you then randomly distributed each bucket into two pools and told one pool they now had to cycle to and from work and the other that they were obliged to travel to work using public transport or by car. If after that you had a 39% reduction in obesity amongst those directed to cycle versus those who had to travel in other ways - that would be real evidence that cycling causes a 39% reduction in obesity.

Taking a sample of people and noting that there is 39% less obesity amongst those who choose to cycle is not the same thing. The obvious explanation for that is that obese people are less likely to choose to cycle.

In the same way noticing that professional "go" players tend to be oriental is not evidence that playing go causes you to become oriental, instead it is evidence that oriental people are more likely to play go.

A correlation shows that maybe it is worth investigating whether there is a causative factor (as it did for smoking) it doesn't establish causation. Especially not when there is a much simpler explanation.

Anyway - let's get out to the sunshine rather than argue statistics - I am heading for a beer in the garden. Which (being a beer drinker) makes me more likely to be obese, but maybe I should conveniently rely on this study and decide I can drink 39% more beer on the basis that the cycle to work will cancel it out

Chuchote

Mm.

"The landmark 1964 U.S. Surgeon General's Report concluded only that 'cigarette smoking is causally related to lung cancer in men'."

Kinda big group was excluded!

RobFowl

Chuchote wrote: »

Mm.

"The landmark 1964 U.S. Surgeon General's Report concluded only that 'cigarette smoking is causally related to lung cancer in men'."

Kinda big group was excluded!

And Doll and Hill's study was on doctors on the medical register !

Theoretically you could argue lung cancer was confined to male doctors

Chuchote

By the way, does anyone know what the intention is with the big wall being built at Beatty's Avenue beside the Dodder? I thought it was just a short section and then the railings will continue, but they seem to be extending it.

If they extend it much further it will turn that lovely pathway by the Dodder into an ideal place for women and boys to be attacked - a narrow place between high walls where no one can hear a scream and where there's no escape.

And what's with the giant rocks they've dotted down along the water there?

tomasrojo

On the meaning of statistical associations, Ian Walker posted this after a clearly frustrating day on Twitter when the BMJ published the study showing lower incidence of cardiovascular disease and cancer among commuting cyclists:

The world seems to have a lot of people who think they're the only one ever to have discovered that correlation doesn't prove causation

https://twitter.com/ianwalker/status/855422231387754497

magicbastarder

so there were a lot of posts pointing out the difference between correlation and causation the day that report was published; but does ian walker understand the difference between correlation and causation, with his conclusions?

tomasrojo

He wrote an excellent introduction to statistical analysis, so I presume he has some strategies for deciding when correlation might mean causation!

magicbastarder

my post was meant to be humorous. i have failed.

tomasrojo

magicbastarder wrote: »

my post was meant to be humorous. i have failed.

Sorry, I just took it as an opportunity to plug his book! I did assume you were being sarky/facetious.

(I did get the joke in the end!)

tomasrojo

Incidentally, Wikipedia has a pretty decent short rundown of ways of ascertaining when correlation might mean causation.
https://en.wikipedia.org/wiki/Correlation_does_not_imply_causation#Use_of_correlation_as_scientific_evidence

Chuchote

The three latest studies on cycling and illness:

https://www.sciencedaily.com/releases/2016/11/161101100149.htm

Recreational, commuter biking linked to lower cardiovascular disease risk
November 1, 2016

People who bike regularly for pleasure or to commute have a lower risk of cardiovascular disease, according to two studies.

In a study of 45,000 Danish adults published in the American Heart Association journal Circulation:

Those who regularly biked for recreation or to commute had 11 percent to 18 percent fewer heart attacks during a 20-year follow-up.
As little as half an hour of biking per week provided some protection against coronary artery disease.
People who took up biking during the first five years of follow-up also had about a 25 percent lower risk of developing heart disease, compared with those who remained non-bikers in the subsequent 15 years.
Researchers estimate that more than 7 percent of 2,892 heart attacks during the 20 years could have been averted by taking up cycling and doing it regularly.

"Because recreational and commuter biking is an easy way to make physical activity part of one's routine in a non-structured and informal fashion, based on the results, public health authorities, governments and employers ought to consider initiatives that promote bicycle riding as a way to support large scale cardiovascular disease prevention efforts," said Kim Blond, M.Sc., lead author and research assistant at the University of Southern Denmark.

In a study published in the Journal of the American Heart Association, middle-aged and older Swedish adults who biked to work were less likely than non-bikers to be obese, have high cholesterol, high blood pressure or pre-diabetes -- all critical drivers of cardiovascular risk.

Researchers followed more than 20,000 people in their 40s, 50s and 60s for 10 years and monitored their commuting habits, weight, cholesterol levels, blood glucose and blood pressure.

At the beginning of the study, compared with passive commuters who used public transportation or drove to work, active commuters who biked to work were:

15 percent less likely to be obese
13 percent less likely have high blood pressure
15 percent less likely to have high cholesterol
12 percent less likely to have pre-diabetes or diabetes
In a follow-up exam 10 years later, those who maintained or took up biking had a 39 percent lower risk of obesity, 11 percent lower risk of high blood pressure, 20 percent lower risk of high cholesterol and 18 percent lower diabetes risk.

and

http://www.gla.ac.uk/news/headline_522765_en.html

Cycling to work may cut your risk of premature death by 40%

Issued: Thu, 20 Apr 2017 00:00:00 BST

Using your bike to get to work could cut your risk of developing cancer and heart disease by almost half.**

New research by the University of Glasgow and published today in the BMJ, has found that cycling to work is associated with a 45% lower risk of developing cancer and a 46% lower risk of heart disease, compared to a non-active commute.

Overall the study found that commuters who cycled were associated with a 41% lower risk of premature death.

Walking to work was associated with 27% lower risk of developing cardiovascular disease and a 36% lower risk of dying from cardiovascular disease, but not cancer or premature death overall.

This study analysed data from 264,337 participants from UK Biobank who were asked questions about their usual mode of commuting to work and then followed up for 5 years. The new cases of cancer, heart attacks and deaths in that 5-year period were assessed and related to their mode of commuting.

magicbastarder

tomasrojo wrote: »

Sorry, I just took it as an opportunity to plug his book! I did assume you were being sarky/facetious.

(I did get the joke in the end!)

colour me disappointed if someone on twitter did not bring this to his attention.

MediaMan

https://www.facebook.com/samhowardS3/posts/10155338783470712

An eloquent piece on the frustration of sharing the road with people who are out to teach cyclists a lesson.

adrian522

Nice piece of driving...

https://twitter.com/CyclingTodayEn/status/860586269667151878

Thargor

The BBC described it as "tormenting a cyclist":

https://twitter.com/BBCSurrey/status/860765975792291840?s=09

manafana

id read some mixed comments on twitter from this, cyclists saying he put himself in danger by not keeping left which is utter hogwash. Van driver showed no patience then decided to not only pass as quick as possible but swipe the cyclist

Chuchote

manafana wrote: »

id read some mixed comments on twitter from this, cyclists saying he put himself in danger by not keeping left which is utter hogwash. Van driver showed no patience then decided to not only pass as quick as possible but swipe the cyclist

Really crazy especially from a professional driver.

Fighting Tao

Chuchote wrote: »

Really crazy especially from a professional driver.

Most van drivers have no more training than the average car driver. I wouldn't necessarily call them professional.

Harrybelafonte

Great to see so many people out on their bikes this morning.

On a related note, it sucks to see so many people out on their bikes this morning.

Lusk_Doyle

Harrybelafonte wrote: »

Great to see so many people out on their bikes this morning.

On a related note, it sucks to see so many people out on their bikes this morning.

Bloody cyclists.

Harrybelafonte

Lusk_Doyle wrote: »

Bloody cyclists.

"Sunday cyclists"

Chuchote

Fighting Tao wrote: »

Most van drivers have no more training than the average car driver. I wouldn't necessarily call them professional.

By definition they're professional, as their profession is driving.

Idleater

Chuchote wrote: »

By definition they're professional, as their profession is driving.

Not in this case I'm guessing. vidette uk is a construction company. If they were a courier company I'd be more inclined to agree with you. Just like having a bike to work ebike doesn't make the rider a professional cyclist.

OleRodrigo

Any elephants around Howth today ?

I'll get my coat.

Chuchote

Idleater wrote: »

Not in this case I'm guessing. vidette uk is a construction company. If they were a courier company I'd be more inclined to agree with you. Just like having a bike to work ebike doesn't make the rider a professional cyclist.

You're right; according to the letter of apology the person driving the van was an engineer.

Which raises the question - should everyone who drives a company van with the company logo on it be trained in defensive/advanced driving? Yes, dar liom.

Idleater

Chuchote wrote: »

Which raises the question - should everyone who drives a company van with the company logo on it be trained in defensive/advanced driving? Yes, dar liom.

I'm a ROSPA qualified driver (motorcycle so rider I guess) and from most of the conversations I've had with my instructors and assessors, the people that gain most from defensive driving courses are those choosing to do it. Once someone is forced to do it, the retained practices vary greater - the idea being, "my employer told me I had to be here to get the cert so I can get the job, what is the minimum I have to do to pass the test" type attitude.

Interestingly, ROSPA advanced tests are only valid for 3 years, full retests are always required, your membership gets you one so it's "free" if you pass. IAM are different and not recognised as an advanced licence in many places outside Ireland iirc.