The computer is not behind a Swedish Pisa-turnaround – financial times

Education. the Results from the latest Pisamätningen is not reliable since the OECD made a blunder, claims economist Gabriel Sahlgren.

But the department for Education do not share his image – Pisa is the most robust samples available.

Which friend showed Sweden improved results in Pisa 2015, and the comments have subsequently been taken in the major.

But Gabriel Heller Sahlgren, phd student at the London School of Economics and affiliated to the Institute for business research, warned in an op-ed in Today’s Society the other day to pull large switches on the result.

Last did the students namely Pisatestet at the computer rather than with paper and pencil. The change may have favored the countries with the greater proportion of IT-based teaching and vice versa. Accordingly, there is a risk that a country’s results in 2012 can not be compared with the result in 2015. To this may be the case is indicated in a comparison between the huvudprovet and a computer-based test sample in mathematics in 2012 where a clear resultatskillnader tagged in several countries, including Sweden.

Not comparable

But Anders Auer, the Agency’s lead author of the Swedish Pisarapporten, says that the two tests in 2012 had separate poängskalor and therefore can not be compared. The digital sample 2012 are not comparable with the 2015 computer-based test. In addition, the OECD for the most recent test done a large pilot study to minimize the differences between paper and screen. In short: the Transition to a digital Pisa has no more than marginal effects on the result.

– have you been able to sort the data that kept the same difficulty level when you brought them over to the computer from the paper. Pisa is the most robust measurement you can get, ” says Anders Auer.

He also says that some countries ‘ decline in income may have quite other explanations.

– it would depend on the test method is only speculation, built on few observations, ” says Anders Auer.

Reversal of the trend, in spite of everything

Gabriel Heller Sahlgren insists, however, that the OECD should have gone more thorough approach and identied the possible consequences more than on the global, international level.

– To go over to the computer without doing a proper analysis of how it affects students in different countries is a blunder.

According to him, it is not possible to exclude the possibility that, for example, Norway’s and Denmark’s rise depends on the digitisation process.

– But it looks like that the Swedish results are reliable.

At that point he is in agreement with Anders Auer: Despite possible uncertainties Sweden got a break in the trend in 2015. The positive turnaround in the Timss tests, which are written with a pen, talking to Swedish students actually made a collection.

the FACTS

Results in the Pisa 2015

Area Score 2015 Points 2012 Change

Science 493 485 +8 (improvement not statistically significant)

reading Comprehension 500 483 +17

Math 494 478 +16

Source: OECD

TT