US Reading Scores Plummet on Global Exam | Hacker News Reader

US Reading Scores Plummet on Global Exam

nytimes.com

41 pointstoomuchtodo17 comments

TIL: B-S-J-Z (China) in the rankings means Beijing, Shanghai, Jiangsu, and Zhejiang.

Instead of testing mainland China as a whole, the OECD allows these four specific provincial-level jurisdictions to represent China in the triennial PISA rankings for 15-year-old students.

Why is China constantly allowed special treatment like this?

I bet if the US only submitted scores from select wealthy Massachusetts and NoVA counties you’d see some crazy numbers.

It will only amount to 3% of the US population. China’s cut out amounts to 13% percent. A 1.41 billion is a lot of people and I understand they don’t want to skew the results. If just taking account that slice, which a quick search says it’s 180 million people, that reading is more than half of the US population. It’s a matter of scale. I’m sure they can add some other provinces to total the US population and still have a higher score.

> China’s cut out amounts to 13% percent

It’s a legally selected sample too: you face restrictions/bans on enrolment in schools there unless permitted. Residing there isn’t enough.

https://en.wikipedia.org/wiki/Hukou

TBH, its EVEN more filtered than that. Zhongkao cuts out "dumb" kids. Migrant kids with wrong houkou can't drag down BSJZ scores, i.e. "dumb" kids get filtered out. But dumb is relative, IIRC PISA does ~10 Chinese provinces, even the shit tier ones that don't make BSJZ cuts had global average PISA 20 years ago, i.e. when those places were legit impoverished. Those provinced developed a lot since, their scores should regress towards tier1 mean - relative to growth a third tier city has more now than 1st tier 20 years ago. The main skew is like 20-30% of actually dumb kinds, because they got diverted to vocational, their education fine but less rigourous.

Eitherway, hard to say, but look at proxy indicators, there's fuckload of "dumb" Chinese international students... i.e. those who couldn't hack GaoKao that are basically A students in western tertiary. And as others have mentioned, BSJZ + kids that pass zhongkao is basically more 2+ standard deviation talent pipeline than OCED combined, disproportionately directed at technical tertiary, i.e if you want to know who has most big brains whose going to be doing the cooking in the future, BSJZ is good indcator.

It will be interesting to see the race-wise results. On 2022, those show US Asians at number 1 for Reading and Science and 4 (behind Singapore, Macau, Taiwan) in Mathematics; with US White close behind; and then a steep drop off for other races until US Black at number 35, 43, 52. There’s a large gap in subpopulations which means change in population blend causes the total population to show change without an underlying intervention to the actual subjects.

But of course we will know that once the race tables are released.

What's your source for the 2022 difference?

PISA itself doesn't publish data broken down by racial subgroup. Back in 2018, the NCES public download included a RACETHC variable (https://nces.ed.gov/pubs2021/data/2021019_rawdata.zip).

But the most recent file on the NCES PISA 'data products' page is from 2018: https://nces.ed.gov/surveys/pisa/publications.asp

(At the time of writing if you expand 'Data Products' the first file is 'Program for the International Student Assessment (PISA) 2018 Public Use File (PUF) NCES Number: 2021019 Release Date: July 8, 2021')

Rahim, you know this data better than me, but this purports to be it. Is it not real or is this somehow subtly different from what I would read it to be? It calls itself the reading results from PISA 2022 by race. You’d have to set a new dependent variable to get the other two. I don’t know how to get it to show all 3 at once.

https://nces.ed.gov/surveys/pisa/idepisa/report.aspx?Lang=10...

If you can’t load it for some reason (that’s a direct link to the report and it launches a query) see here: https://imgur.com/a/gOnTJQZ

Thanks, Roshan. I didn't know this interface existed.

I had thought the only way to access the data was via downloadable data files (no longer published) or the PDF reports.

I'm curious what the trend for whites in the US is over time. That would be a good first approximation on how much of this decline is due to culture or the educational system (including COVID disruptions) and how much is demographic change/immigration.

It would be interesting to see how US born black people's scores differ from black immigrants from Africa who tend to value education much more.

Somewhat related but I saw this tweet that claims the SAT and ACT has been significantly relaxed to produce higher scores. The competition is to try to make an easier test so more people use that test and pay the fees. The post was from someone selling an alternative, so it should be taken w a grain of salt. But if you look at Harvard, the median SAT score is 1550 today, much higher than before.

Another post gave an example of a real SAT question, which I will present below:

The following text is from David Barclay Moore’s 2022 novel Holler of the Fireflies. The narrator has just arrived at summer camp, which is far away from his home.

This place was different than I thought it would be. I’d never been somewhere like this before. I did feel scared, but also excited.

According to the text, how does the narrator feel about being at summer camp?

A) He feels overjoyed.

B) He feels peaceful.

C) He feels both scared and excited.

D) He feels both angry and jealous.

https://x.com/the_culturist_/status/2092961940314144888/phot...

>Another post gave an example of a real SAT question, which I will present below:

As other X users have noted, it might just be a "easy" question and not reflective of the entire test. Also from a test perspective it makes sense to have "easy" questions, so the scores have proper scaling. If you're designing a test to judge science ability, but all the questions are about quantum physics, then you'll have a bunch of people scoring 0 just because it's outside of their ability and have no hope of answering.

Aktually scoring a zero would mean they all knew something about QM

https://archive.today/nKJbQ

This is a friendly reminder that for about 10 years, "sightwords" (rote memorization) displaced phonics in an attempt to show off a younger reading age.

(Plot twist: You can get kindergarteners to "read" via sightwords, when they might otherwise only be intellectually-mature enough to do so (for real) starting in 1st-2nd grade... but the shortcuts catch up with folks when they hit ~4th-5th grade reading comprehension).

Here is a link to the actual report which is better than this paywalled editorialization.

https://www.oecd.org/en/publications/pisa-2025-results-volum...

Reply on news.ycombinator.com