Basic science and statistical understanding

[MrsTerryPratchett]

I'm not a scientist. I do have a BSc. but nothing more impressive. I do wince when I hear some assumptions that people make who don't have basic science (and statistical) knowledge. What basic scientific or statistical principles do you wish the general public (and members thereof) knew about?

I'll start. Causation and correlation. A correlation doesn't prove causation. Particularly when that correlation is a correlation of one. "I had the flu vaccine and got the flu really badly" isn't causation.

When I’m teaching causation v correlation l like these graphs
tylervigen.com/old-version.html

Also don’t keep telling me how many people died today without telling me how many people usually die on a Sunday mid November as number of dead without context is meaningless.

[quote Cathpot]When I’m teaching causation v correlation l like these graphs
tylervigen.com/old-version.html[/quote]
I just used this on on another thread...

Unrelated to covid (sorry but I feel like you may be my people here!), but people claiming that low average life expectancy in the past is the same as saying that no one lived above 40. Argh! Average life expectancy figures often skewed by huge infant and child mortality FFS!

Used to be a teacher and even saw that one in textbooks. Drives me nutty.

@Cathpot

Also don’t keep telling me how many people died today without telling me how many people usually die on a Sunday mid November as number of dead without context is meaningless.

This is the biggie for me

Along with the general lack of interrogation

@Gertrudetheadelie

Unrelated to covid (sorry but I feel like you may be my people here!), but people claiming that low average life expectancy in the past is the same as saying that no one lived above 40. Argh! Average life expectancy figures often skewed by huge infant and child mortality FFS!

Used to be a teacher and even saw that one in textbooks. Drives me nutty.

A good place for my mean, median and mode rant! Yes, life expectancy estimates are skewed depending on which average you choose.

However, I have liked some of the reporting on this recently. With quotes like, "only 3% of the population of Congo is over 65, whereas 25% of the population of Canada is over 65" [numbers made up by author ] because that actually gives you a pictures of the actual ages, not those skewed by infant mortality.

@MrsTerryPratchett

I'm not a scientist. I do have a BSc. but nothing more impressive. I do wince when I hear some assumptions that people make who don't have basic science (and statistical) knowledge. What basic scientific or statistical principles do you wish the general public (and members thereof) knew about?

I'll start. Causation and correlation. A correlation doesn't prove causation. Particularly when that correlation is a correlation of one. "I had the flu vaccine and got the flu really badly" isn't causation.

My favourite example of correlation v causation is getting sunburn while wearing sunglasses.

That's true!

My favourite example of correlation v causation is getting sunburn while wearing sunglasses

Excellent example of confounding though!

This is of my favourite David Spiegenhalter quotes. It’s from his book “The Art of Statistics.” (Or so Twitter told me.)

"I am often asked why people tend to find probability a difficult and unintuitive idea, and I reply that, after forty years researching and teaching in this area, I have finally concluded that it is because probability really is a difficult and unintuitive idea."

I teach ice cream sales correlating with deaths by drowning therefore we should all be wary of ice cream around water.

MrsTerryPrachett have you seen this?

I often wonder how this got commissioned but it’s pleasingly quirky

@Changechangychange

My favourite example of correlation v causation is getting sunburn while wearing sunglasses

Excellent example of confounding though!

Also the perennial "it's hot outside, I must put on sunscreen". Always drives me nuts...

Just had to explain on another forum that the data being skewed towards older people doesn't mean it was fixed.

Also that 50,000 isn't half a million.

I've come on this thread for an antidote.....

People often give greater weight to personal anecdotes and their own google research than peer reviewed scientific data. I agree with the poster who said that people lack awareness of their own limits of expertise. This is worrying but also understandable given the breakdown in trust between the public and politicians and scientists.

I wish people would understand that exponential is not a synonym for big. Growth can be exponential now, but still low. However if it continues to grow exponentially then it will not stay that way.

Also R is coming down means things are good. No. If R stays above 1 you will still get exponential growth, things will just be fucked less fast. R getting lower does not imply cases coming down.

Not strictly science/stats, but related:

That CMO etc "only cares about graphs". No, he's simply doing his job!!!

That the scientists "bullied" BJ into the lockdown. That would only be true if BJ wasn't doing his job, and, in a different way, the scientists weren't doing their jobs.

Basic measurement.

Comparing any covid data to data gathered in March is utterly utterly pointless because the data gathered in March only told you how many people with severe symptoms tested positive in hospital. There is practically no data at all about how many mild or asymptomatic infections were present. Yet, I saw that Financial Times of all fucking idiots claiming that lower deaths now mean that the infection fatality rate must have gone down. They looked at the data and rather than saying 'a comparison isn't possible due to poor data,' they drew the most ridiculous and least likely conclusion.

I would also love to grab the next person who says 'exponential growth' and force them to explain to me in excruciating detail how the modelling that predicts exponential growth works. Because I can fucking guarantee with 100% certainty that they haven't the first clue. They don't know that the model is pure guesswork, contains zero actual data, has been proven wrong on a very large number of assumptions and is essentially a piece of total bollocks. No, they just say 'exponential growth' over and over as though it's fact. Fucks me off like nothing else.

This thread has been ...cathartic.
And it has also reminded me of some areas where I think I 'get it' whereas I really, really don't. This is a very good thing!

Average life expectancy figures often skewed by huge infant and child mortality FFS!

And maternal.

When I was in Y10 studying the industrial revolution, I read something which said that in 1840s Merthyr the average life expectancy was about 17. I couldn't my teenage head around how people would have had enough times to have got jobs, got married and had children by the age of 17.

Duh.

Thank you for this thread. It's interesting.

@MrsTerryPratchett

And that is a COVID thing because people keep saying, "the average age of people who die from COVID is over the average age people generally die" like that's an indication that we shouldn't care. That very much depends on how those numbers are calculated.

That oft-repeated covid average death age thing makes me do a headdesk so hard I'm in danger of getting splinters.

I love the dancing statistics!

Good way to teach it.

Scientist here. Issues that annoy me are the lack of understanding of the following: Causation and correlation, statistical significance, sample sizes, bias, the difference between a review and peer-reviewed research. And don’t get me started on all the PCR experts out there 😡

Maybe we should start a list of terms that people bandy aroud:

PCR
RNA
Exponential
R0

Any I've missed?