PISA sampling does not explain Iceland’s poor results, professor says

An associate professor at the University of Iceland says it is a persistent myth that other countries handpick students for PISA assessments, while Iceland includes almost all 15-year-olds. mbl.is reported that Berglind Gísladóttir, an associate professor at the university’s School of Education and one of the authors of Iceland’s national PISA report, rejects the claim.

The claim has resurfaced on social media alongside an image comparing the estimated number of 15-year-old students in several countries with the number who took the test. The image states that 78 per cent of Icelandic 15-year-olds participated, compared with much smaller percentages in most other countries.

Some people have questioned whether the results are reliable when such a large proportion of Iceland’s age group takes part. Others have suggested that students in other countries are selected in a way that produces better results. The origin of the image is unknown.

“This is a very persistent myth that comes up every single time: that other countries select their students to participate, while everyone has to take part here. But that is not how it works,” Berglind told mbl.is.

She said Iceland is below the minimum number of students required for PISA, which is why all eligible students are included. Ninety-nine per cent of schools took part, along with about 83 per cent of students in the relevant PISA age group, she said.

About 17 per cent of students are exempted from taking part for various reasons, including learning difficulties or an exemption from attending school. Berglind said the proportion is very similar in all participating countries.

According to Berglind, countries must follow strict rules when selecting samples. The process involves a type of cluster sampling, which she said is difficult to explain simply.

“This follows very strict scientific conditions, and if countries deviate from the rules set by the OECD for any reason, they are marked with an asterisk and excluded from the comparison,” she said.

She said the way students are selected therefore has no effect on the results or on performance. Iceland’s weak results cannot be explained by its sample, she said. About one in four students who took the test last year did not reach basic proficiency in reading comprehension, science or mathematics, and no country has declined as much as Iceland since 2015.

“The selection is carried out in such a way that it should be fully representative of the student population in that country. If people had studied methodology, they would know that a sample of 1,000 from a population of 300,000 provides almost the same degree of accuracy as a sample of 1,000 from a population of 300 million,” Berglind said.