Those aged 16–24 in 2012: 11 years on, they read 27 points better in Finland and 39 worse in Poland
One generation in two rounds of the test: in 2023 they were 25–34. It reads significantly better in 7 of 27 countries and worse in 6; elsewhere the change is within the error.
More
The adult test does not follow the same people, but one birth cohort can be compared: the young of 2012 and people born in the same years in 2023. Study, work and reading after 20 keep changing the skill.
Each bar is the 2023 mean of 25–34-year-olds minus the 2012 mean of 16–24-year-olds: the four largest rises and four largest falls, all beyond the error.
- Finland: 16–24 in 2012 → 25–34 in 2023 points
296.7 → 323.8 (+27.1 ± 8.6)
- Denmark: 16–24 in 2012 → 25–34 in 2023 points
276.1 → 296.9 (+20.8 ± 7.9)
- England: 16–24 in 2012 → 25–34 in 2023 points
265.4 → 285.3 (+19.9 ± 9.1)
- Estonia: 16–24 in 2012 → 25–34 in 2023 points
287.1 → 303.5 (+16.4 ± 7.5)
- South Korea: 16–24 in 2012 → 25–34 in 2023 points
292.9 → 273.4 (−19.5 ± 8.3)
- Slovakia: 16–24 in 2012 → 25–34 in 2023 points
276 → 252.1 (−23.9 ± 8.2)
- Lithuania: 16–24 in 2012 → 25–34 in 2023 points
278.4 → 250.1 (−28.2 ± 8.4)
- Poland: 16–24 in 2012 → 25–34 in 2023 points
281.5 → 242.9 (−38.6 ± 7.8)
Macroscope's calculation from table A.3.7 of the OECD 2023 report: the 2012 score = the 2023 trend score minus the published change. Error from the published standard errors and the 3.27 linking error; significant beyond 1.96 errors.
The age windows do not match exactly: 16–24 in 2012 is 27–35 in 2023, not 25–34. Migration and deaths change the cohort. The OECD does not publish this calculation.
A question for this picture
Where did the share of weak adult readers rise most?