The structure of the survey was similar to that used previously. A field test took place in early 2006 and the main survey in spring/summer 2007. In England and Scotland, pupils in 150 to 160 schools at each age group were randomly selected to represent their country. TIMSS sets high participation targets, in order to ensure that data is of good quality and that valid comparisons between countries can be made.

In both age groups, sampled pupils completed a single test booklet containing items assessing achievement in mathematics and science. The item formats varied and included closed-response items, short-answer items and more extended items focussing on problem-solving and investigatory approaches in mathematics and science.

Findings from TIMSS are detailed. For 2007, they include:

• Achievement data, by subject and by content domain within each subject
• Achievement data by cognitive domain within each subject (knowing, applying, reasoning)
• Comparative data, showing each country's achievement against that of other countries, by subject and by content and cognitive domains
• Trend data, showing each country's achievement over time, by subject and by content and cognitive domains
• 'Benchmark' data, describing the achievements of pupils in different bands, and the proportion of pupils in each achievement band, for each country
• Contextual information, at country level, school level and pupil level
• Analysis of the interaction between contextual variables and achievement

International findings from TIMSS 2007 and previous cycles are available on the International Study Center website: http://isc.bc.edu.

The national report for England for TIMSS 2007 can be accessed here

The national report for England for TIMSS 2003 can be accessed here

The TIMSS 2007 national report for Scotland can be accessed at
http://www.scotland.gov.uk/Publications/2008/12/08150124.

Participating schools were randomly selected by the international team which runs the survey. The international team also specified which class(es) and pupils should participate. Participating countries are not allowed to invite other schools if the selected schools decline, and schools not drawn in the sample cannot opt to participate. This approach ensures that each country's sample is robust and comparable with that of other participating countries.

	Number of participants		Percentage of sample participating (weighted)
	Schools	Pupils	Schools	Pupils
England year 5	143	4316	90	98
England year 9	137	4025	86	88
Scotland P5	139	3929	94	94
Scotland S2	129	4070	86	90

We are grateful to the many schools, teachers and pupils who gave up their time to represent England and Scotland in this important study.