Five things you probably dont know from PISA

Five things you probably don’t know from PISA….

1. Shanghai did 50 points worse on the computer test in 2012 than in the paper test…. .

2. Why does Viet Nam do so well? Country Singapore Japan Country Finland Macao Canada Country China South Korea New Zealand Slovenia England Mean Country 556* Estonia 538* Taiwan Mean Country 531* Vietnam 529* Hong Kong 528* Mean 518 516 513 512 Country Australia Germany Netherlands Switzerland Ireland Mean 534* 532* Mean 525* 523* Mean 510 509 506 503*

2. Why does Viet Nam do so well? Only 50% of its 15 -year-olds are eligible to take part! 650 75 th percentile including pupils outside target population 600 Figures refer to the 75 th percentile in science. 550 576 519 ‘Before’ = reported results 500 450 400 75 th percentile excluding pupils outside target population 350 850 ‘After’ = including 15 -year-olds who are not in school, assuming they would all perform below the national median.

3. Around 1 in 3 American schools declines to participate…. .

4. You can only really compare PISA reading results back to 2009 (not 2000)…. . “Changes in design and construct coverage were particularly important in earlier PISA assessments. The change in performance observed between PISA 2000 and later assessments may thus not always reflect genuine changes in what students know and can do……… …. . the associated uncertainty associated with comparisons involving PISA 2000, 2003 and 2006 reading results with later results is only imperfectly captured by linking errors. …. some caution is needed when interpreting reading trends before PISA 2009. Very few link items 3 questions all from 1 unit – which was not technically the best unit……

30 Original 5. A lot of things changed in PISA 2015 from previous cycles. . . …. and this had a non-trivial impact upon some of the results Albania 20 “The negative changes between PISA 2012 and PISA 2015 reported for Chinese Taipei (-18 score points) and Viet Nam (-17 score points) are, to a large extent, due to the use of a different scaling approach. Had the PISA 2012 results for mathematics been scaled with the PISA 2015 calibration sample and the PISA 2015 approach to scaling, the differences in results for Chinese Taipei and Viet Nam would have been only -3 points and -4 points, respectively” Uruguay 10 0 Singapore -10 Viet Nam Chinese Taipei -20 -30 Turkey -30 -10 10 Re-scaled 30