Hacker News Reader

Comment by username90 | Hacker News Reader

KTH tests engineering students math abilities by giving new students the same test every year. There is nothing gained or lost from this test so nobody studied for it. Anyway, this gives us a reasonable trend in how well students know the basics, and it shows that students are getting better and not worse right now. The worst class was the one starting 2007.

https://www.kth.se/aktuellt/nyheter/trenden-har-vant-matteku...

View on HN

Note that KTH is an elite institution so there is a selection bias. It's like saying education in America must be going well because new MIT students are getting better.

Now you are exaggerating. KTH is of course a good school for Swedish standards but calling it elite is weird. In what way would you say that KTH is elite, because I genuinely do not understand how the word "elite" applies to this school?

I'm a bit out of touch, but of all schools in Sweden, the only I'd call elite - by the common meaning of the word - is Operahögskolan (I think it's called, the Opera school) and possibly Stockholm School of Economics and the medical schools of the larger older universities + Karolinska. Those places are highly competitive.

Everyone I know, from the little village I grew up in Sweden, who wanted to go to KTH, did so.

Maybe elite is the wrong term, I'm not a native speaker... what I meant is that KTH is probably the best technological university in Sweden and one of the best tech unis in Europe overall, so the math level of its admitted students is not necessarily representative of the general level.

It is the best technological university in Sweden, but it's also one of the biggest universities, if not the biggest. It accepts a lot more than the elite, in any sense of that word.

> who wanted to go to KTH

Here's your selection bias. Self-selection is also selection.

I'm not sure what you are claiming here. Are you saying that I'm wrong in questioning calling KTH "elite" because I used one anecdote in a larger argument?

You can look up the admission statistics yourself: it's very easy to get in to KTH. Maybe there are some programs that are more competitive than others.

When I finished high school, the admission criteria for example Electrical Engineering was roughly "you need a non-failing grade in all high school courses". How is that "elite"? Compare with medical school where it used to be "you need the best grade in every single course otherwise you're not getting in". Or the Opera school where there's an admission test in front of an audience.

I can guarantee that MIT is more difficult to get into than KTH.

When I entered KTH you needed top grades in all subjects to be admitted to engineering physics. Electrical engineering was probably not even the third hardest to enter.

You are drawing conclusions about Swedish education based on a biased sample (people admitted to KTH).

It's true that looking at people admitted to elite medical schools would introduce even bigger bias, but it does not mean that the sample you chose is representative.

No I'm not drawing any conclusions about Swedish education, I'm talking about KTH specificially. Are you replying to the wrong person?

You need to read the whole thread for context. The original post claims education quality in Sweden deteriorated. username90 argues it didn't, using unchanged KTH test results as support. JorgeGT says KTH is an elite institution so there is a selection bias. You disagree, saying that KTH is not an elite institution. I'm saying the selection bias is still present because KTH students are not a representative sample.

In this context it does not matter how elite KTH is. It only matters whether KTH students on average are materially different from general population.

Thanks. I contacted the author of that article and he gave me an updated graph. It, of course, doesn't have the numbers for 2020, as the test wasn't taken because of COVID-19.

Den mån 15 feb. 2021 kl 07:55 skrev Hans Thunberg:

Hej Kim! Förkunskapsprovet genomfördes även vid utbildningstart i augusti 2017, 2018 och 2019. I höstas blev det ingenting p ga av pandemien (vilket ju verkligen är synd – man hade vilja veta hur distansundervisning som började redan våren -20 påverjade förkunskaperna)

De förbättrade resultaten från 2015 och 2016 höll väsentligen i sig de tre följande åren, här är ett diagram som visar lösningsfrekvensen på förkunskapsprovet hos antagna till civilingenjörsutbildningarna vid KTH 1997 – 2019.

Med vänliga hälsningar Hans Thunberg

https://drive.google.com/file/d/1V7VrCVAAFia0Rd4gLRYI0dfEIZU...

So they have already been accepted... Not a truly random sample.

“Nobody studies for it” is surely an overstatement. Do not underestimate the fear of tests.

Nobody studies for it since nobody knows they will take the test. It is marked as a blank spot on their calendar and people gets surprised when a test gets handed out. Then they quickly forget they ever took a test since so much happens in the first days of new students.

The only acceptance criteria is grades, or a local SAT/ACT like test but without essays or other harder to grade parts. It's language, logic and math. See more here [0]. You don't get in because you're good at sports, it's not even a consideration.

I've taken a similar test at another university in Sweden and truly nobody studies for it, in the introduction week it was simply done at one point as part of the introduction curriculum, without even being mentioned beforehand.

So, it's a fairly consistent sample of people aiming for a STEM education.

[0]: https://en.wikipedia.org/wiki/Swedish_Scholastic_Aptitude_Te...

Reply on news.ycombinator.com