I despair of the education system.

takman

paddyrg wrote: »

I know enough statistics to be generally untrusting of any figures anyone uses to try to sell me a product or ideology, the sample sizes used in shampoo adverts in particular are insulting. And as for trial methodology...

I've actually seen adverts that base their claims for their products on less than 50% of a small sample "agreeing" that it worked. But yet because some of the people agreed they can then make that claim on the advert! Lol

NBLondon

paddyrg wrote: »

I believe wholeheartedly in double blind randomised trials where the trial is registered before beginning (so the results cannot be retrospectively used to demonstrate a different conjecture, or 'bad' trials buried until random chance throws up a 'good' one). I cringe on a regular basis when I read 'x cures/causes cancer' on the front page of the Mail, when the researchers behind it would immediately and completely refute that extrapolation of a very marginal set of results.

Have you read Bad Pharma by Ben Goldacre? Explains exactly how this has been got away with for many years... Given that doctors (usually intelligent individuals) are often fooled, no wonder the average person doesn't question.


The classic one is relative versus absolute improvements. A new drug reduces the chances of the patient dying from X from 2% to 1%. Is that a reduction of 1% or of 50%? Guess which one the drug company will tell you? If the new drug is 5 times as expensive as the alternative - is it a good deal?


But when many people struggle with maths (or refuse to engage with maths) they don't ask that question.

Nick_C

NBLondon wrote: »

The classic one is relative versus absolute improvements. A new drug reduces the chances of the patient dying from X from 2% to 1%. Is that a reduction of 1% or of 50%?

If 100 people out of 10,000 are dying each year instead of 200, then it's a 50% reduction in the number of deaths. The chances of dying from the disease are reduced from 2% to 1%. Both figures are true and equally valid.

Of course, someone who can't multiply 4 by 3 is unlikely to understand this data.

bod1467

As the old expression goes ... There are lies, damned lies, and statistics.

paddyrg

Nick_C wrote: »

Well, it is or it isn't.

A+!A = 1, right?

(A.!A = !(A+!A)), wouldn't you say?

jackieblack

takman wrote: »

I've actually seen adverts that base their claims for their products on less than 50% of a small sample "agreeing" that it worked. But yet because some of the people agreed they can then make that claim on the advert! Lol

I've known people 'pass' exams which have a passmark of less than 50% :huh: :wall:

Pennywise

jackieblack wrote: »

I've known people 'pass' exams which have a passmark of less than 50% :huh: :wall:

In last Summer's higher level GCSE Maths exam, you'd get a C grade with a score of under 35%! (33 or 34% I think it was). So get two thirds of the paper wrong and you still get what they class as a "good pass" - because of course all grades are allegedly passes these days in this "no one can fail" world. I remember the days when anything under 50% was a fail, i.e. more wrong than right.

GwylimT

Pennywise wrote: »

In last Summer's higher level GCSE Maths exam, you'd get a C grade with a score of under 35%! (33 or 34% I think it was). So get two thirds of the paper wrong and you still get what they class as a "good pass" - because of course all grades are allegedly passes these days in this "no one can fail" world. I remember the days when anything under 50% was a fail, i.e. more wrong than right.

Thats because the C grade questions are around 40% (maximum) of the marks available. So if someone can answer a majority of the C grade questions correctly, but not the B, A or A*, what do you think they should get?

If you sat the foundation paper to get a C grade you would need around 63% for a low C grade

GwylimT

jackieblack wrote: »

I've known people 'pass' exams which have a passmark of less than 50% :huh: :wall:

Universities have a passmark of 40%.

Pennywise

takman wrote: »

But even if I had to work out the area I could still easily do it using my phone and get the exact answer without mental arithmetic required!.

Why would you even need the phone?

Carpets come in fixed widths, typically 3 or 4 metres as shown on the back of the sample and you buy length usually by the metre. (Though some shops do half metre).

So if the extremes between each side of your room is 3.5 metres and the extremes of it's length is 4.5 metres, you round up and you need a carpet of 4 metres wide by 5 metres long, 4 times 5 (simple times tables as per primary school) gives 20 square metres.

Basic primary school maths of rounding up and simple times tables. The basic maths skills aren't the issue. It's finding out the width and length options from the shop and also being able to measure the extremes of your room properly. Even with a calculator, it's garbage in, garbage out, if you don't get the fundamentals right.