3 ms·
EF's dataset is students who take EF tests (eg. TOEFL). What this degradation means is that TOEFL scores are much lower. This makes sense because there's less
by alephnerd 2y ago
EF's dataset is students who take EF tests (eg. TOEFL). What this degradation means is that TOEFL scores are much lower.
This makes sense because there's less of a need to travel abroad for studies, and it has also become much more difficult for Chinese nationals to study abroad over the past 4 years (eg. Zero COVID from 2020-2023, Visa restrictions in the US).
Edit: The TOEFL's methodology has also changed to actually test for speaking and writing fluency. Before, the TOEFL would overindex reading skills, but not speaking and writing.
- kelipso 2y agoAlways know where the data comes from! I see so many examples of people generalizing to the whole population of a country when the data comes from a specific subset.
- yorwba 2y agoEF does not administer the TOEFL, but a competing English test, the EF SET, which can be taken online for free. When fewer people take a test, I would expect the average score to go up, since the people who do take it are likely more motivated. But it's all within the range of historical variation anyway. https://www.ef.com/wwen/epi/regions/asia/china/ https://www.ef.com/wwen/epi/regions/asia/china/
- alephnerd 2y agoDoh! Good callout. Confused ETS with EF! And this is even a good point as well - the EF SET isn't really used in most cases where English fluency is critical (eg. College admissions in the West, Citizenship/PR applications in much of the Commonwealth, etc), so the kinds of students taking EF specific tests might be less prepared or knowledgeable about English and Western culture compared to those preparing for TOEFL or IELTS.