The rise and fall of achievement
Although overall, Scotland's international ranking for the three key subject areas has fallen since the last survey in 2003, it continues to perform above the OECD average.
However, some countries are doing better and at a faster pace, according to the latest Programme for International Student Assessment, published this week.
Bob Kibble, senior lecturer in science education at Edinburgh University and chair-elect of the Association for Science Education Scotland, said it would not be fair to judge Scottish science education solely on the Pisa survey.
Nevertheless, the results showed evidence of healthy attainment in the 101 Scottish secondaries which took part. Findings of low interest in pursuing science as a long-term career mirrored other research, including the ROSE (Relevance of Science Education) survey, he said. There were early signs that A Curriculum for Excellence would make science more human and relevant and less fact-based, which made him optimistic for its future, said Mr Kibble.
The Educational Institute of Scotland welcomed the country's "solid performance" overall, but warned that continuing investment was needed to close the attainment gap between the highest and lowest achievers.
Ronnie Smith, EIS general secretary, called on the Scottish Government to learn lessons from other successful comprehensive systems, such as Finland's - "a highly inclusive system which has benefited from significant government investment in highly professional teachers and top-quality resources over the years".
Maureen Watt, the Skills and Schools Minister, said the findings represented an inherited challenge which the new Scottish Government was determined to tackle. She said the study showed that the impact of a disadvantaged background had a greater effect on how Scottish students performed than in many other countries.
In 2003, when the last survey of 15-year-olds was carried out, three countries performed significantly better than Scotland in science. In 2006, four countries performed better in science, eight in maths, and five in reading.
The latest survey focused on science, where Canada, Finland, Japan and New Zealand scored significantly higher than Scotland - although the results this year may be skewed as the test framework was changed from previous surveys.
In maths, Scotland came behind the same four countries as well as Australia, Korea, the Netherlands and Switzerland. However, its mean score dropped by 18 points compared with 2003 - the largest drop in scores of all the OECD countries.
In reading, Canada, Finland, Ireland, Korea and New Zealand all had significantly higher mean scores than Scotland.
Scotland's performance in the upper levels of attainment for maths and reading fell significantly. In maths, the total percentage of students scoring at Level 3 and above (the scale goes up to Level 6), fell from 69 per cent in 2003 to 60 per cent in 2006.
In reading, 60 per cent of students scored at Level 3 or above in 2006 compared with 68 per cent in 2003.
The survey showed that the vast majority of Scottish students exceed basic levels of attainment in science; at higher levels, they also exceed OECD averages. Only Finland has significantly more students scoring above Scotland at the second top level, Level 5. However, at Level 4, Scotland was outperformed by a number of countries.
Gender did not make a significant difference in science. But in maths, gender differences have become significant, with boys now recording a 16-point advantage over girls (compared with the OECD average of 11). Boys' performance dropped by 14 points and girls' by 23 points. For reading, the gap is 26 points, in favour of girls, and is thus significant, but largely unchanged from 2003.
Scottish students showed a relatively low level of interest in science - in common with other high-performing countries such as Australia, Finland and the Netherlands. Paradoxically, students in some of the lowest-performing countries such as Turkey, Portugal and Mexico, showed the highest level of interest.
An expert who is highly critical of Scotland's education standards relative to other countries has issued a health warning about how seriously international surveys should be taken as a measure of performance.
In an analysis this year of a range of international surveys, including PISA and PIRLS (see opposite), James Stanfield, of Newcastle University's School of Education, said Scotland's results were "a mixed bag".
This was not surprising as none of the organisations used comparable methods or samples; samples in individual countries were small and varied; they often reflected, not the performance of the governments, but policies of several years previously.
The PISA survey is designed to measure the attributes of children of a certain age rather than the performance of schools, Mr Stanfield stressed in an analysis which he carried out for the Policy Institute, a right-leaning think tank.
Scotland's PISA results include pupils from independent schools, which may improve the overall ranking. Unlike countries such as Germany, Scotland did not include pupils from its special schools. Thus, on a like-for-like basis, Scotland's results would appear much poorer.