Zompist Bboard Again

Torco wrote: ↑Fri Dec 20, 2024 8:59 pmand yeah, third world is often used like this, no? to mean poor country?

It is.

Starbeam wrote: ↑Fri Dec 20, 2024 11:32 am I understand why it would seem like a good idea to use a small sample of the population to reflect a greater whole, but in practice this always just leads to stereotypes and incomplete data.

It seems to me that we use stereotypes and incomplete data for almost everything. What's the alternative?

Torco wrote: ↑Fri Dec 20, 2024 8:59 pm okay sure, it'd be better to perform a census of the whole population, but you have any idea how expensive those are? nah, for most purposes picking people as randomly as you can is the way to go.

I get you can't survey everyone, but there's a difference between "less than 1%" and "the entirety of the group". Yes, I'm aware representatives help a lot for things, but i do feel uncomfortable reducing humanity to, like, a republic of unique groupthinks. Not the most coherent thought, but I hope you get the idea. Again, it would help if they were all very upfront with the limits of their survey.

i don't know what panels operate in the us, but like, there's a lot of programs to answer surveys get points you can redeem for home appliances. lemme check... na, apparently netquest works by invitation only ? [and that's a good thing for pollers, cause you want your sample to be more or less representative: the more self-selected it is, the more biased your results] but like... this ? https://www.indeed.com/career-advice/fi ... s-for-jobs

It's a start! Thanks, Torco

my dude, i'm well off by chile standards and i make less than us minimum wage. and chile is not that poor, have you looked at living conditions in peru? argentina has literal 50% poverty these days. check conditions in bolivia, in venezuela, in costa rica, in haiti. this is very much the third world, i assure you. and check the PISA scores, they're not great.

and yeah, third world is often used like this, no? to mean poor country?

I'll take your word for it. I always perceive it to mean things even worse, but I won't overspeak if you know they're that bad.

rotting bones wrote:It seems to me that we use stereotypes and incomplete data for almost everything. What's the alternative?

Not using stereotypes, at least not consciously. Is there some beneficial use to stereotypes I need to know?

We all use incomplete data, but there's being limited and not even trying. Demographic stuff I understand for marketers and political pollsters, but I'm less supportive it coming along for scientific research.

Starbeam wrote: ↑Sat Dec 21, 2024 2:34 pm

We all use incomplete data, but there's being limited and not even trying. Demographic stuff I understand for marketers and political pollsters, but I'm less supportive it coming along for scientific research.

What about medical research? Should new drugs be tested in experiments involving large shares of the entire population before being approved?

Raphael wrote: ↑Sat Dec 21, 2024 2:43 pm

Starbeam wrote: ↑Sat Dec 21, 2024 2:34 pm

We all use incomplete data, but there's being limited and not even trying. Demographic stuff I understand for marketers and political pollsters, but I'm less supportive it coming along for scientific research.

What about medical research? Should new drugs be tested in experiments involving large shares of the entire population before being approved?

That one makes sense. I think biological differences in humans are a lot easier to handle than social ones. They probably already have some chemical precendence, too. But again, the more the merrier.

your point is not absurd, but mathematically, it ends up being very similar to have a sample of 1k and 10k people. you only start getting significant gains at the double digits percentage of people, and that's WAAAY too many people, WWWAY too expensive.

Starbeam wrote: ↑Sat Dec 21, 2024 2:34 pm Not using stereotypes, at least not consciously. Is there some beneficial use to stereotypes I need to know?

We all use incomplete data, but there's being limited and not even trying. Demographic stuff I understand for marketers and political pollsters, but I'm less supportive it coming along for scientific research.

Why shouldn't we always try to use representative data? Using all the data we can get might make the results unrepresentative because it doesn't reflect the actual distribution of demographics in society.

maybe the point is that sex-age-income-region of residence are probably not the only things we should be controlling for in research of a scientific character? and like, yeah, there's probably all sorts of other confounding variables that might be at play. but then again, you can't control for 33 different things. okay, you can, but it'll cost you. everything'll cost you

kikknos wrote: ↑Tue Dec 10, 2024 6:10 pmHow do you feel about where your nation placed, what do you think accounts for its standing and how would you improve the results either for your country or the others?

I'm surprised that Sweden placed so high in both literacy and numeracy. Lately there's been a lot of talk about young people's lack of reading comprehension. Supposively a lot of the new students starting university have too low reading comprehension for university level studies. I heard one problem (of many) is that students don't have the endurance to read a long text, because nowadays they may use audiobooks for reading assignments. Then again, the results of this survey do show the younger age group doing a little worse than the next age group.

For a longer time now there's been a lot of talk about math skills getting poorer. Students are getting good grades in math (for the whole subject in junior high or high school), but not in the national tests (a standardized test everyone in the country has to take at the same time). And the PISA ranking is getting worse too.

I'm also surprised about Singapore, South Korea and Israel being so low down in this survey. The stereotype is that people in these countries study really well, isn't it?