NAPLAN 2019: Reasons to be cheerful

By Jennifer Buckingham

There were some dramatic headlines when the preliminary results from the 2019 NAPLAN tests were released earlier this week.

“NAPLAN: $20bn flop, schools fail to lift most students’ academic results” – The Australian
“Writing wrongs: Our kids worse off 10 years after start of controversial school tests” – The Daily Telegraph
“NAPLAN results show we are failing our children” – The Courier Mail
“NAPLAN changes the way we raise our children, and not for the better” – The Sydney Morning Herald.

How accurate are these headlines?

Putting aside the silly suggestion that NAPLAN is negatively affecting parenting or, as was suggested elsewhere, is somehow to blame for the poor results, it depends on a number of things.
1.    The NAPLAN domain (reading, language conventions, writing, numeracy)
2.    The time period of analysis
3.    The school year group in question

I will not attempt to parse all of the possible permutations of these factors, but I will look more closely at primary school reading as an example of how national results do not tell us enough to be able to draw conclusions about achievement trends and their possible implications.

Australia does not have a national education system. It has six states and two territories, each of which has its own public education system, its own Catholic system and a variety of independent schools, some of which are autonomous and some of which are in small systems. They all have different policies and practices. So while national trends are important to follow, we can’t confidently trace a line between any national policy and NAPLAN trends. The possible exception is school funding, but since the impact of funding is mediated by the way in which the funding is spent, even this is tenuous without further analysis.

The lowest level of systemic analysis in the public reporting on NAPLAN is at the state and territory level. The NAPLAN data published by ACARA do not break down results in a way that allows comparison between government and non-government systems and sectors.

When we look at the state level results for reading in Year 3 and 5, the picture is more optimistic. There have been incremental improvements in all states and territories since NAPLAN testing began in 2008. In some states, the improvements have been significant and deserve to be acknowledged.

Queensland started from a low base. It had the second lowest mean score for Year 3 reading in 2008. Its mean score is now not significantly different from Western Australia, Tasmania and South Australia. Part of this improvement is likely to be attributable to the fact that the age requirements for enrolment in the prep (Foundation) year of school changed in 2015, and became compulsory in 2017. Prior to that many children in Queensland had had fewer years of school before taking the NAPLAN tests. This is now consistent with other states and territories.

Victoria started as one of the highest performing states in Year 3 reading and has also improved. Its mean Year 3 reading scores are the highest in the country. However, there is a potential contributing factor to consider – participation rates. While Victoria has edged out New South Wales in terms of mean scores (445 for Victoria versus 435 in NSW), it had a markedly lower participation rate (94.8% in Victoria versus 98.1% in NSW). New South Wales and Western Australia are the only states where participation rates increased from 2008 to 2019, with NSW increasing by a higher margin than WA.

It is difficult to know whether or how the Victorian students who did not participate in NAPLAN might have affected the mean score for the state but it is a factor to consider.

While mean scores are important, some of the greatest improvements have been in the proportion of children who did not achieve the national minimum standard (NMS). Queensland again showed one of the greatest improvements but there were also significant decreases in New South Wales, Western Australia, South Australia and the Northern Territory.

While the national minimum standard is a low benchmark – and there is a good argument that a higher and more meaningful proficiency benchmark should be determined – it is still important to note that in some states, the proportion of children below the NMS has reduced by half or more.

In NSW, WA, SA and the NT, there have been substantial shifts in early reading instruction policies and practices towards more explicit and systematic methods, particularly for phonics. It is highly likely that this has contributed to the improvement, especially given research and case studies showing that these reading instruction practices are a characteristic of high performing primary schools. For example, here, here, and here.

The improvement in Year 5 reading scores is not as great. This is to be expected, given that it takes time for improvements in the early years of school to flow through to later years.

‘Matthew effects’ also come into play in later years; achievement gaps get larger and more difficult to ameliorate as children progress through school because the factors affecting reading achievement become more complex and cumulative in nature. This is especially apparent in the reading results for Years 7 and 9. Students who have mastered decoding and accurate word reading can have their reading comprehension constrained by limited vocabulary and background knowledge.

Nonetheless it is important to acknowledge that there has been improvement in NAPLAN reading scores in Year 3 since its inception; to fail to do so hazards creating a sense of futility and malaise. People in government departments and non-government system authorities, and principals and teachers in hundreds of schools around the country, have been working extremely hard, sometimes at the risk of their careers, to implement evidence-based reading instruction in an effort to raise literacy levels. The latest NAPLAN results indicate that it has not been in vain.