The cognitive costs of air pollution • Hacker News Board

Re: The cognitive costs of air pollution

#41

Post by gnode » Mon, Nov 18, 2019, 5:34 PM UTC

glofish wrote: post #12

larnmar wrote: post #3

If the effect were present, and as large as implied, you’d expect low IQ test results in ultra-polluted Chinese cities such as Shanghai and Beijing. And yet... http://www.city-data.com/forum/world/2348902-china-iq-map-pr...

No that is not what it is claimed. Pollution reduces your IQ, not that it makes your IQ lower than someone else's. The people living in those cities form a so-called "self-selecting" sample. They are probably, on average better educated, and they recognize the value of education much better. They already would score higher than the others, that's why they moved to the city. It might just as well be that the same peop…

The original post doesn't mention IQ. IQ is a terrible metric for studying short-term cognitive effects.

Re: The cognitive costs of air pollution

#42

Post by rbavocadotree » Mon, Nov 18, 2019, 5:34 PM UTC

From the original source:

> Stock market returns are lower on polluted days.

I don't want to dismiss the entire article, but I think this points to possibility that there is some bad statistics going on here. I don't see how daily stock returns are proof of, or even related to, cognitive ability on a daily basis.

This is exactly the type of finding you'd expect with p-hacking.

Re: The cognitive costs of air pollution

#43

Post by learc83 » Mon, Nov 18, 2019, 5:35 PM UTC

zionic wrote: post #19

learc83 wrote: post #6

Earlier quoted context omitted.

>IQ tests standardized? IQ tests can't be standardized across cultures. Even something like Raven's Progressive Matrices has a cultural component. https://www.apa.org/monitor/feb03/intelligence Also China has some issues with the way they report scholastic achievement that make their results appear higher than they should. China is very selective about which students take standardized tests that are used for global c…

Just because a test might have some component you can vaguely articulate as having some cultural influence does not make the test itself invalid for comparison across cultures. Claims that basic pattern recognition is somehow not a valid means of assessment across cultures are dubious at best.

>does not make the test itself invalid for comparison across cultures

If you can't accurately quantify the cultural component, then yes it does.

>Claims that basic pattern recognition is somehow not a valid means of assessment across cultures are dubious at best.

Basic pattern recognition isn't the only thing that's being measured--that's the problem. Even something as simple as how often a student has been exposed to a matrix of rows and columns has an impact on performance. Repeated exposure to similar tests also increases performance.

Re: The cognitive costs of air pollution

#44

Post by PeterStuer » Mon, Nov 18, 2019, 5:39 PM UTC

So does this mean we can now expect populist politicians to crank up PM2.5 emitters at a massive scale prior to elections?

Re: The cognitive costs of air pollution

#45

Post by bearforcenine » Mon, Nov 18, 2019, 5:44 PM UTC

graeme wrote: post #35

rahuldottech wrote: post #17

This really bothers me. I live in a VERY polluted place, and unfortunately moving away won't be an option for me for at least half a decade. What can I do to minimise the health effects of such an atmosphere? I'm planning on getting an indoor air filter for my home.

Starts with monitoring. Awair or laseregg are good quality monitors. Then you can assess how well filters work. I think there are masks you can use outside too. You can also check which times of day/days are worst. And which streets. You can carry the laseregg around for local measures.

PurpleAir also makes an indoor air quality sensor that measures particulate matter: https://www2.purpleair.com/products/purpleair-pa-i-indoor

Not sure how it compares with Awair or laseregg.

I have no affiliation with PurpleAir, but they are pretty popular in Utah.

Re: The cognitive costs of air pollution

#46

Post by olivermarks » Mon, Nov 18, 2019, 5:45 PM UTC

So logically living and working in the countryside is optimum for your IQ rather than polluted cities and areas?

Re: The cognitive costs of air pollution

#47

Post by LessDmesg » Mon, Nov 18, 2019, 5:46 PM UTC

Pollution, IQ, who cares?!? We've got carbon dioxide to fight, a benign gas that for some reason we believe is the root of all evil! Climate change is the most important issue of our time, not pollution (which is actually real and not forged by IPCC, but who cares about reality when there is Greta Thunberg)

Re: The cognitive costs of air pollution

#48

Post by lazyjones » Mon, Nov 18, 2019, 5:51 PM UTC

rbavocadotree wrote: post #42

From the original source: > Stock market returns are lower on polluted days. I don't want to dismiss the entire article, but I think this points to possibility that there is some bad statistics going on here. I don't see how daily stock returns are proof of, or even related to, cognitive ability on a daily basis. This is exactly the type of finding you'd expect with p-hacking.

This is probably just a coincidence. Stock markets perform lower on Mondays: https://www.macroption.com/stock-market-performance-volatili...

Traffic is probably generally higher on Mondays, though I have only anecdotal evidence.

Re: The cognitive costs of air pollution

#49

Post by 9dl » Mon, Nov 18, 2019, 5:52 PM UTC

graeme wrote: post #35

rahuldottech wrote: post #17

This really bothers me. I live in a VERY polluted place, and unfortunately moving away won't be an option for me for at least half a decade. What can I do to minimise the health effects of such an atmosphere? I'm planning on getting an indoor air filter for my home.

Starts with monitoring. Awair or laseregg are good quality monitors. Then you can assess how well filters work. I think there are masks you can use outside too. You can also check which times of day/days are worst. And which streets. You can carry the laseregg around for local measures.

Just google brands and find that they may be not so accurate or may be calibrated for some specific region of China

https://smartairfilters.com/en/blog/how-accurate-are-common-...

So may be better solution are local laboratories with calibrated equipment

Masks are tricky.

Most "light" masks are for huge particles and can't defend against most of dangerous chemicals

Military and science masks are working good against some chemicals and do not protect from another. Except solutions with O2 tanks on your back

Re: The cognitive costs of air pollution

#50

Post by ajross » Mon, Nov 18, 2019, 5:53 PM UTC

rbavocadotree wrote: post #42

From the original source: > Stock market returns are lower on polluted days. I don't want to dismiss the entire article, but I think this points to possibility that there is some bad statistics going on here. I don't see how daily stock returns are proof of, or even related to, cognitive ability on a daily basis. This is exactly the type of finding you'd expect with p-hacking.

It's a review article. The paper you mention was just one of many cited. And sure, those are all different papers written by different authors and it's surely not unlikely that there are errors in there. But the reason for reviews like this one is to point out that all these different results support the same basic hypothesis.

As far as stocks, specificaly: really? People with cognitive trouble make bad stock picks all the time. Most of us have grandparents who have exhibited exactly this kind of mistake. While it's surely true that the bulk of analyst-driven trades are checked by methods that aren't sensitive to pollution (or just by analysts in different climates), there are certainly enough single-decider trades going on to show a small effect like this if it exists.

I mean, no, one oddball stock market paper doesn't prove much of anything. But in combination with a bunch of other research like this, it's worth taking seriously.