9 out 18.5 cats agree

[agric]

In the good old days, statistics produced to sell a product comprised
8 out of 10 cats prefer....



or "in a survey of 100 women 18 percent agreed this shampoo was the best"
so why lately are apparently random numbers of people questioned 35 out of 141 people asked, said they preffered this skin product



confused....

[MiniMan64]

Because they had to discard the answers of the other 859 people who thought the product was vile.

[wiffmaster]

Probably because they sent out 2000 questionnaires and only got 141 back.

[TheEnd]

random numbers, long words that sound almost like science and "our tests" are believed by a lot of people.

of both sexes.

usually the sex that buys shampoo with nano-lipisoids and a soothing hydrospakled ceradabrilator.

[agric]

cant be that simple,

they'd just rephrase the questions and poll another few hundred surely?

No I was thinking it was some statistical formula to somehow massage an otherwise dull outcome

[W00DY]

The weirdest one was for a butter last year where upon reading the stats at the bottom you'd be more likely to either like the other product or not be fussed than prefer it but because it was slightly more popular than the comparison product on it's own they made a big fuss about it being the best.

[Morningside]

agric said:

cant be that simple,

they'd just rephrase the questions and poll another few hundred surely?

No I was thinking it was some statistical formula to somehow massage an otherwise dull outcome

15% out of 100 people hated it. Keep asking and asking until the figures rise and come to the answer

27% out of 1023 people loved our product.


Added: Much worse are cosmetic products that sell 10,000 products a year and yet they ask less than 100 people and STILL come out with lowish figures.

Edited by Morningside on Tuesday 23 February 22:11

[Fort Jefferson]

I heard quite a frightening statistic on the radio today.

That 1 in 16 people will get some form of bowel cancer.

[fadeaway]

depends if it's a scientific study, a survey or just an indication of how many people feel a particular way.

Scientific study will usually end up being out of a non-round number (eg 141), due to number of people who respond, or due to people in the study group being discounted/dropping out for various reasons.

A general survey can just literally ask 100 people, and wouldn't need to discount anyone.

If it's just a stat used to give an indication, then I guess the number will usually be used as a ratio or percentage, rather than quoting the actual raw figures.

[LuS1fer]

This intrigued me and so I polled 97 people and 37.576% of them didn't care while the remainder went out and bought a statistics and marketing book.

[Frankeh]

Internet polls..

[Oakey]

Next time you see one of those L'Oreal adverts on TV, keep an eye on the figures at the bottom. They change. One minute it says 73% of 180 women agree, then for reasons unknown it changes to 83% of 180women agree.

[jbudgie]

Fort Jefferson said:

I heard quite a frightening statistic on the radio today.

That 1 in 16 people will get some form of bowel cancer.

It's about 1 in 18 for men and 1 in 20 for women.

[Frankeh]

I thought it was 1 in 3 people get cancer at some point.
Well, it said 1 in 3 get affected by it.

That could mean directly getting it, family getting it or friends getting it.

Or it could mean 1 in 3 people get cancer.