Students in London do not replicate their success at GCSE when it comes to the Programme for International Student Assessment (Pisa), new analysis has found.

While London teenagers achieve some of the highest GCSE results in the country, this is not mirrored in their Pisa scores, according to research by John Jerrim, professor of education and social statistics at the UCL Institute of Education.

Professor Jerrim calculated that, in both 2009 and 2012, students in London performed less well in Pisa than those in other parts of the UK.

If you combine data from both years, London’s students had a mean score of 479 in maths compared with 495 for the rest of the UK - a difference of 16 points. 

Pisa: Which pupils are the world’s biggest ‘bullshitters’?

Related: England’s pupils among the most segregated by ability

News: ‘It is ridiculous that Pisa is now making teachers anxious’ 

On average, students in London also scored 16 points less than the whole of the UK in reading, while the gap was widest in science - at 19 points on average. London was also below the Organisation for Economic Cooperation and Development (OECD) average in all areas. And in maths, London’s students were three years behind 15-year-olds in Shanghai.

London tends to outperform the rest of the UK when it comes to GCSEs  - this year, its GCSE performance was the second-best in the UK after Northern Ireland, with 25.7 per cent of students achieving top grades of 7 or above. So Professor Jerrim said it was surprising that this was not replicated in Pisa data.

In a blog post summarising his findings, he wrote: “One obvious question to ask, then, is why is there no evidence of a “London effect” in Pisa?”

Professor Jerrim looked at whether London’s high ethnic diversity could provide any explanation for the disparity. He analysed data from the National Pupil Database, and found that “pupils from disadvantaged and ethnic minority backgrounds in England were likely to do much worse on the Pisa test than one would anticipate, given how they performed in their GCSEs”. 

For example, black and Asian teenagers scored almost 30 Pisa points lower than predicted, given their GCSE scores. 

Professor Jerrim concluded that “more work needs to be done to understand why pupils from ethnic minority and disadvantaged backgrounds perform so differently across these two assessments.”

The regional differences were less pronounced in the more recent 2015 Pisa rankings. “Yet [London] was still some way behind the top-performing area,” Professor Jerrim said.