PISA 2018: Australian students’ performanceResearch 3 Dec 2019 10 minute read
A long-term national decline in reading, maths and science achievement has led, for the first time, to Australia’s achievement failing to exceed the OECD average in one of the assessment domains.
The Australian Council for Educational Research (ACER) released the first volume of national reports for the 2018 Programme for International Student Assessment (PISA) in December 2019.
PISA assesses the reading, mathematical and scientific literacy of 15-year-old students around the globe. Conducted every three years since 2000, it enables countries to measure and compare how well prepared students nearing the end of compulsory schooling are to use their knowledge and skills to meet real-life challenges. Each cycle focuses on a major domain, allowing students extra time to complete this section. In 2018, reading was the focus.
More than 600 000 students in 79 countries and economies took part in PISA 2018, including a nationally representative sample of 14 273 Australian students in 740 schools.
Understanding the results
PISA results are reported on a set of scales that enable comparisons between average scores as well as between proficiency levels, which describe the skills and knowledge that students demonstrate. The word ‘significant’ is used to indicate that the difference between two results is statistically significant and unlikely to have arisen by chance. It is not to be confused with the term ‘substantial’.
Due to Australia’s large number of participating students, which are spread across multiple year levels, it is possible to estimate the number of score points associated with one year of schooling. For Australia, one year of schooling corresponds to around 33 points on the PISA reading literacy scale, around 28 points on the mathematical literacy scale, and around 27 points on the scientific literacy scale.
Australian achievement in decline
Australia’s average achievement in reading, maths and science has experienced a pattern of long-term decline. The decline is such that the average achievement of an Australian 15-year-old in 2018 is: almost one year of schooling behind in reading compared to an Australian 15-year-old in 2000; more than one full year of schooling behind in maths compared to in 2003; and almost one full year of schooling behind in science compared to in 2006.
Figure 1: Australian achievement in PISA since 2000, measured from the first cycle in which a subject was the major focus domain
Prior to the 2018 assessment, Australia’s achievement on PISA has always been above the OECD average in each of the three domains. In 2018, however, Australia’s mathematical literacy performance was similar to the OECD average.
According to Dr Sue Thomson, ACER Deputy CEO (Research) and National Project Manager for PISA in Australia, the only country whose maths performance has fallen further than Australia’s is Finland – although they still outperform Australia.
Five countries whose mathematics performance was on par with Australia’s in their first PISA assessment now outperform Australia; and, of 16 countries whose maths performance was lower than Australia’s in their first PISA assessment, nine now outperform Australia and seven are now on par with Australia.
Comparing PISA 2018 results internationally, Australia performed significantly below 10 other countries and economies in reading, 23 in mathematics and 12 in science, after accounting for insignificant differences between the scores of countries and economies.
Figure 2: Performance of all countries/economies with available data in relation to Australia
Number of countries/economies
* Based on available data (77 countries for reading and 78 for maths and science).
The highest performer in PISA 2018, in all domains, was the grouped Chinese provinces of Beijing–Shanghai–Jiangsu–Zhejiang (B-S-J-Z China), followed by Singapore. By comparison to B-S-J-Z China, Australian students performed at a level roughly one-and-a-half school years lower in reading, around three-and-a-half school years lower in maths, and around three years lower in science. Compared to the highest performing country, Singapore, Australian students performed at a level one-and-one-third school years lower in reading, around three years lower in maths, and around one-and-three-quarter school years lower in science.
Australia’s participation in PISA is part of the National Assessment Program. As such, the Measurement Framework for Schooling in Australia identifies a proficiency at or above Level 3 in PISA as the agreed National Proficient Standard, as it ‘represents a “challenging but reasonable” expectation of student achievement’ at that age.
Between 2015 and 2018, the proportion of Australian students achieving the National Proficient Standard in reading and maths did not change significantly. In science, however, the proportion of Australian students who attained the National Proficient Standard declined by three percentage points.
Figure 3: Percentage of students achieving the National Proficient Standard
The proportion of Australian students achieving the National Proficient Standard was slightly above the OECD average in reading (59% in Australia) and science (58% in Australia), but the same as the OECD average in maths (54%). In B-S-J-Z China, around 90 per cent of students reached this proficiency level in maths and science, as did 80 per cent in reading. Even if Australia only took results from its top performing region in each domain, the Australian Capital Territory, the proportion achieving the National Proficient Standard (70% in reading, 66% in maths and 71% in science) would still be well below that of B-S-J-Z China.
Results within Australia
In reading, students in the Australian Capital Territory, Western Australia, Victoria, Queensland and South Australia performed at a higher level than the OECD average, while students in New South Wales, the Northern Territory and Tasmania performed on par with the OECD average. Victoria, Queensland and the Northern Territory were the only jurisdictions not to have recorded a decline in average reading achievement since the domain was first focused on, in 2000. The only jurisdiction to record a significant change in reading literacy performance since PISA 2015 was the Australian Capital Territory, where it increased by 19 points or more than half a year of schooling.
In mathematics, the average performance of students in the Australian Capital Territory and Western Australia was higher than the OECD average. Students in Victoria, Queensland and New South Wales performed on par with the OECD average, and students in South Australia, Tasmania and the Northern Territory performed at a level lower than the OECD average. Average performance in mathematics has declined in all Australian states and territories since the domain was first focused on, in 2003. There were no significant changes in the average maths scores of each jurisdiction since the 2015 cycle of PISA.
In science, students in the Australian Capital Territory, Western Australia, Victoria, Queensland, South Australia and New South Wales performed at a higher level than the OECD average, while students in Tasmania and the Northern Territory performed at a similar level to the OECD average. Victoria and the Northern Territory were the only jurisdictions not to have recorded a decline in average science achievement since the domain was first focused on, in 2006. The only jurisdictions to record significant change in scientific literacy performance since PISA 2015 were the Australian Capital Territory, where it increased by 6 points, and New South Wales, where it decreased by 12 points.
ACER’s report compares the achievement of Australian students in terms of their background characteristics, such as gender, socioeconomic background, Indigenous status, immigrant background and language spoken at home.
In terms of gender, girls outperformed boys in reading by the equivalent of around one year of schooling, which was similar to the difference recorded in 2015. In science, the small difference in achievement in favour of boys recorded in 2015 was no longer present. In mathematics, however, a small but significant difference in favour of boys has returned since the closure of the gender gap in 2015.
Average performance in reading, maths and science has declined in the long-term across all socioeconomic (SES) quartiles, with the largest declines recorded among the highest quartile. As in previous cycles of PISA, the difference in achievement between students in the highest SES quartile and students in the lowest SES quartile is equivalent to around three years of schooling.
Non-Indigenous students continue to outperform their Indigenous peers by the equivalent of more than two years of schooling but, while non-Indigenous student performance has declined in all domains over the long-term, Indigenous student performance has not changed significantly.
Across the various cultural groups assessed, average performance has declined in the long term in reading, maths and science, except for foreign-born students in reading.
In terms of immigrant background, first-generation students performed at a higher level in reading than both Australian-born and foreign-born students, who performed at a similar level. In mathematics, the performance of first-generation and foreign-born students was similar and above that of Australian-born students. In science, first-generation students performed at a higher level than foreign-born students, while the performance of Australia-born students was not significantly different to first-generation or foreign-born students.
In terms of language background, students who spoke English at home performed at a higher level in reading and science than students who spoke a language other than English at home. In mathematics, there was no difference in performance between these groups.
ACER’s report also compares the performance of the three Australian school sectors: government, Catholic and independent. Government schools are the only sector not to record a decline in reading performance since 2009, the year that Australian school sector performance was first reported. In maths and science, average performance since 2009 has declined across all school sectors.
In all three assessment domains, students in independent schools performed higher than students in Catholic schools by the equivalent of almost one year of schooling, and students in Catholic schools performed higher than students in government schools by the equivalent of around three-quarters of a year of schooling.
However, because there are higher proportions of students from lower socioeconomic backgrounds who attend government schools compared to the proportions of students from lower socioeconomic backgrounds who attend Catholic or independent schools, to ensure fair comparisons school sector results are adjusted for differences in an individual student’s family background or socioeconomic background as well as the school-level socioeconomic background.
After adjusting for the socioeconomic background at both student and school-level, there were no differences in reading or science achievement between the school sectors. In maths, however, once student- and school-level socioeconomic background were accounted for, there was a difference in performance between government and Catholic schools, where students attending government schools achieved at a higher level. This is the first time in PISA’s history that this has occurred in Australia.
PISA is managed in Australia by ACER on behalf of the Organisation for Economic Co-operation and Development (OECD). Australia’s participation in PISA is funded by the federal, and state and territory governments. Volume II of the PISA 2018 results, due for release in March 2020, will focus on student and school characteristics. To find out more about Australia's participation in PISA, visit www.acer.org/ozpisa
Read the full reports:
PISA 2018 In Brief I: Student performance by Sue Thomson, Lisa De Bortoli, Catherine Underwood and Marina Schmid, ACER (2019).
PISA 2018: Reporting Australia’s Results. Volume I: Student performance by Sue Thomson, Lisa De Bortoli, Catherine Underwood and Marina Schmid, ACER (2019).