A new report by the Australian Council for Educational Research provides a very useful overview of the results for Australian students on international and national tests. In particular, it shows an overall decline in reading and mathematics achievement amongst 15 year-old students, a growing gap between the most advantaged and the least advantaged secondary schools in Australia, and further confirmation of the strong influence of socio-economic background on student achievement.

While the report found a small overall decline in reading and mathematics results on the Programme for International Student Assessments (PISA), it found that it was most pronounced amongst high achieving students. Although this finding is a minor part of the report, it was highlighted in some media reports.

The finding has given weight to those who claim that the emphasis being given to improving the results of students at the lower levels and ‘closing the gap’ is denying opportunities for the top students. However, the report’s finding is too strong because the percentage declines are similar across all levels of achievement and scores for the highest achieving students recovered somewhat since 2006 while those of the lowest achieving students have continued to decline.

ACER’s report failed to identify a distinct difference in the trends of scores for the highest and lowest achieving students since 2000. The reading results of the highest achieving students declined between 2000 and 2006 but increased a little in 2009 while their mathematics scores declined between 2003 and 2006 but did not change in 2009. In contrast, the results for the lowest achieving students declined between 2003 and 2006 and continued downwards in 2009, especially in mathematics in which there was a significant decline.

As a result, there was no reduction in the very large achievement gaps between the highest and lowest achieving students in reading between 2000 and 2009 and in mathematics between 2003 and 2009.

These gaps are strongly associated with the socio-economic status (SES) backgrounds of students. The ACER report shows that the influence of students’ socioeconomic backgrounds on achievement has increased and that there is a growing gap between the most advantaged and the least advantaged schools in Australia.

The report provides additional evidence for changing the way schools are funded so as to provide additional funding for schools and students most in need. This is the essence of the Gonski funding model. It must be implemented nationally. It will provide continuing support for high achieving students but it also provides a once in a lifetime opportunity to reduce disadvantage in education and reduce the gap between rich and poor.

Further details of the trends in the PISA test results for high and low achieving students are as follows.

**Reading**

The PISA results for 15 year-old students show a decline in reading of 13 points from 528 in 2000 to 515 in 2009. Australia was only one of 16 OECD countries whose reading score fell over this period out of 26 countries participating in both PISA cycles. Others included Canada, Czech Republic, Finland, France, New Zealand, Sweden, and the USA, but Australia did have one of the larger declines.

However, the decline was very small in percentage terms at only two per cent and the large part of it occurred between 2003 and 2009 when there was a decline of 10 points.

The decline occurred at all achievement levels (see Chart 1). The absolute point declines were slightly larger at the highest achievement levels (17-18 points) than at the lower levels (8-10 points). However, the percentage declines were similar across the board. The percentage declines at the 95th, 90th, 10th and 5th percentiles were virtually all the same at around 2.5 to 2.8 per cent.

The point score declines were statistically significant at the highest achievement levels; that is, we can be reasonably confident that the declines were not a chance outcome. The point declines at the lowest levels were not statistically significant, so we cannot be certain that they were real declines.

The pattern of change over the period for the highest and lowest achieving students was quite different. Scores for the highest achieving students declined between 2000 and 2006 but have increased since then. The score at the 95th percentile in 2009 was 12 points higher than in 2006 and 10 points higher for the 90th percentile. In contrast, the scores for low achieving students were stable between 2000 and 2003 and have declined since then, with declines of 10-14 points to 2009. Thus, there has been some recovery in the reading results for the highest achieving students since 2006 while those of the lowest achieving students have continued to decline.

There was little change in the achievement gaps between the highest and lowest achieving students between 2000 and 2009 [Chart 3]. There was a significant decline in the gap between the 95th and 5th percentiles and between the 90th and 10th percentiles between 2000 and 2006, but these declines were almost completely reversed between 2006 and 2009.

The gap of 325 points between the 95th and 5th percentiles is equivalent to nearly 10 years of schooling and the gap of 254 points between the 90th and 10th percentiles represents nearly 8 years of schooling.

**Mathematics**

The decline in mathematics was 10 points from 524 in 2003 to 514 in 2009, or just 2 per cent. Australia was only one of 15 OECD countries whose mathematics score fell over this period out of 28 countries participating in both PISA cycles. Others included Belgium, Czech Republic, Denmark, France, New Zealand and Sweden. Several countries (Belgium, Czech Republic, France, Ireland and Sweden) had larger declines than Australia.

The decline occurred at all achievement levels (see Chart 2). The absolute point declines were only very slightly larger at the highest achievement levels (11-12 points) than at the lower levels (7-9 points). The percentage declines were very similar across the board at just below 2 per cent. The declines were statistically significant at the 25th, 75th and 90th percentiles but not at the 5th, 10th and 95th percentiles.

As in the case of reading, the pattern of change was quite different at the highest and lowest levels of achievement. The decline in scores of the highest achieving students occurred between 2003 and 2006 but there was virtually no change in scores between 2006 and 2009. In contrast, the scores of the lowest achieving students increased between 2003 and 2006 and fell significantly between 2006 and 2009. The 10th percentile score fell by 14 points and 18 points at the 5th percentile.

There was little change in the achievement gaps between the highest and lowest achieving students in mathematics between 2003 and 2009 [Chart 4]. There was a significant decline in the gap between the 95th and 5th percentiles and between the 90th and 10th percentiles between 2003 and 2006, but these declines were almost completely reversed between 2006 and 2009.

The gap of 308 points between the 95th and 5th percentiles is equivalent to over 7 years of schooling and the gap of 242 points between the 90th and 10th percentiles represents about 6 years of schooling.