Evidence-based education needs standardised assessment

The latest Gonski review aims to improve evidence-based decision-making in Australian school education. It recognises that governments’ educational investment must be based on more than politics, just as teachers’ instructional decisions must be based on more than intuition. Like other professional sectors, Australian education must be guided by rigorous evidence of what works, for whom and in what contexts.

Standardised assessments, like NAPLAN, are powerful tools in building a strong evidence base for education policy and practice. As NAPLAN enters its second decade, it is timely to reflect on how Australia can make best use of standardised assessment to drive system improvement. This does not deny valid criticisms of punitive standardised testing regimes. Instead, it considers how we might avoid a “baby and bathwater” scenario, and retain the benefits of standardised testing with fewer flaws.

Comparison not competition

Comparison of standardised assessments across systems, schools and classrooms can guide evidence-based policy and practice in many ways. Analysis of NAPLAN trends can help identify policies and practices that may have contributed to improvements. The first Gonski review used comparisons of NAPLAN data as evidence to estimate the costs of quality school education.

Australia participates in international standardised tests like PISA, TIMSS and PIRLS. This is part of a broader global conversation about how to make education systems work better for everyone. Many teachers and school leaders are now using standardised test data to guide school improvement.

On the other hand, standardised assessment can fuel unhealthy competition. The worst effects of MySchool can be seen in residualised schools abandoned by students and families who can afford to go elsewhere. The worst effects of NAPLAN itself can be seen in students placed under pressure to gain the score they need to get into a selective school, or top-stream class.

Internationally, simplistic PISA league tables risk undermining the global improvement agenda that the assessment was designed to support.

Standardised testing does not have to be used this way. It is most effective when used for system improvement, not sanctions or exclusion. Australia has not followed other nations in linking assessment to sanctions for schools or pay for teachers. This is something to be celebrated and sustained.

Standardised not homogenised

Standardised assessments work best when they adapt to students’ individuality. For example, through “tailored testing” in NAPLAN online.

There is potential for Australia to go much further by assessing students across the full continuum of learning, instead of bundling them into year-level groups.

ACER is also developing standardised assessments that use a wider range of methods to capture the skills of students who may not perform their best on a written test. This makes standardised tests more inclusive of different learning styles and cultures, as well as disability.

Assessment for teaching

Standardised assessments can inform what teachers teach, based on evidence of student learning. This happens most effectively when assessments are mapped to curriculum. More work needs to be done to strengthen the connection between curriculum and assessment in Australia. This would help teachers make better use of NAPLAN results to inform their teaching. Current work on describing national learning progressions in literacy and numeracy will help connect the Australian Curriculum to NAPLAN assessment.

We also need to assess the right things. Australia’s National Assessment Program covers a broad range of subject areas, beyond literacy and numeracy. Research is also underway about assessing general capabilities, such as critical and creative thinking, and collaboration, which are essential to students’ success in modern workplaces.

Pluralism not hegemony

A healthy education system will have multiple assessments (large-scale and small), each designed to suit the purpose at hand. NAPLAN is an imperfect measure by nature, and cannot be expected to measure children’s learning as competently as the teacher who spends hours with them every day.

On the other hand, individual teachers’ judgements cannot map learning across the entire education system. Teachers may be experts on the progress of their students, but they cannot compare that progress with students in the school down the road, let alone a school in another state or territory. Standardised assessment provides the best birds-eye view of where the system is working, and where additional attention is required.

Most importantly, standardised assessment is part of the social contract between governments and populations, to provide a quality education for every child.

ACER works with many countries developing standardised assessments, hungry for information about how well their system is working. In countries where government investment is limited, standardised assessments have even been developed by citizen-led groups to meet parents’ demands for information about their children’s learning. This is the best illustration of the purpose of standardised assessment: as evidence that empowers education stakeholders to focus on positive change.