Evidence is critical in education. Topical new research findings related to child and adolescent development, learning processes, education inequities, and the outcomes of specific pedagogical and classroom approaches must be shared with teachers, leaders, and policymakers. Alongside these, stakeholders need contextual information about the quality and nature of that evidence: have findings been replicated with different learners and in different contexts? Are the interpretations drawn by the researchers valid reflections of the data, or are other interpretations possible? How robust are the phenomena? 

Education is multidisciplinary and should draw on evidence from the multiple fields that inform ecologies of learning and teaching. Cognitive and psychological sciences are vital for understanding the psychological foundations of learning, for example, but do not inform our understanding of educational and social systems and their impact. Evidence-based practices must weave together insights from different fields in a way that is rigorous and robust. 

Some evidence is widely replicated and universally applicable (e.g. matching pedagogy to so-called student “learning styles” does not work – and may in fact penalise learning), whereas other evidence may be relevant only in particular contexts (e.g. worked examples are effective in supporting problem solving in Maths, particularly for well-defined problems and novice learners, but are less relevant in English). Supporting teachers, leaders, and policymakers to know what phenomena are universal, or not, and why, is vital – as are discussions about what evidence relates to which aspects of development and learning. 

Critiques of evidence-based practice in education

Evidence-Based Practices have been criticised in EduResearch Matters recently on the grounds that they are harmful and oppressive. These critiques raise important questions regarding the slimness of evidence used by some policymakers; the peculiar interests of some advocates in oversimplifying particular research findings and excluding others; and the focus on experimental interventions to the exclusion of other useful methodologies that can offer different types of insights about education, students, and learning. 

Such critiques can, though, be misread by stakeholders as suggesting that evidence itself is unimportant. This mischaracterisation is unhelpful. Instead, we must be clear that evidence matters – as does the robustness of that evidence; its generalisability or specificity; its ecological validity; and the contextualisation of that evidence for teachers and teaching. 

Those with limited expertise in educational research, including policy-makers, should turn to educational researchers with genuine expertise in specific domains to understand what research shows as our “best-bets”. That is, the pedagogical practices shown to best achieve specific educational outcomes in specific contexts; the degree to which prior knowledge, discipline, age, social context, and learner characteristics affect these bets; and the background knowledge about learners, social contexts, and development that is needed to support other related aspects of schooling such as wellbeing and classroom behaviour.  

The role of universities in promoting evidence-based practice

A key justification for the positioning of teacher education in universities is the need to connect school practice with research scholarship to enhance student learning. As outlined by Aspland (2006), however, this has not always been the case. In the 1800s, school-based apprenticeship models were widely used in Australia. While some conservative commentators prefer this model still, concerns emerged that instructional skills among trainees were poor. In the 1900s, teacher colleges focused predominantly on the craft of teaching. It was not until the late 1980s, and following moves by minister Dawkins to amalgamate colleges of advanced education with universities, that teacher educators in Australia came to adopt more scholarly and theoretical approaches connecting research evidence from different disciplines to teacher education and practice. 

Most academics in Schools of Education today have both teaching and research roles, and there is very little peer-reviewed research in education in Australia that does not come from a university faculty. Key research insights related to cognitive load, worked examples, expertise, reading science, goal setting, neuromyths, formative and summative assessment, EAL/D learners, Aboriginal and Torres Strait Islander history and cultures, and multi-tiered systems of support (MTSS) – all prominent components of the Teacher Education Expert Panel’s new core content – come directly from expert researchers located in Schools of Education.  

Emphasis on peer review

Note here our emphasis on independent peer-review: a universal gold standard in research accountability and quality. TEQSA notes that research, at a minimum, must (i) lead to and/or transmit new knowledge or advances in creative or professional practice in a field, (ii) be a planned, purposive intellectual inquiry, and (iii) produce outputs that are subject to external, independent scrutiny. The Australian Code for the Responsible Conduct of Research similarly states that research must be transparent and must be tested through peer review. 

Of course external organisations may also conduct relevant educational research provided they adhere to the Australian Code. The Australian Council for Educational Research (ACER) is one such example, with an NHMRC-registered Human Research Ethics Committee and peer-reviewed research subject to external oversight and scrutiny. 

The Australian Education Research Organisation (AERO) recommend that teachers have high confidence in an approach only after having “read research published in peer-reviewed academic journals OR conducted by a trustworthy source such as AERO”. This highlights a dilemma for policymakers, school leaders, and teachers seeking accessible evidence not restricted to refereed journal articles – how do such stakeholders confidently differentiate the quality of the range of excerpts, explainers, popular press and commercial tools available and promoted to them without underlying peer-reviewed evidence? And how can university researchers with expertise in ensuring quality support them?  

What should evidence look like in practice? 

To support teachers, leaders, and policy-makers to engage in evidence-based practices, we highlight three important caveats. 

First, context matters. A key role for teachers and school leaders in the translation of research evidence into practice is in knowing how different research insights will apply within their local context. As educational researchers we must be explicit in highlighting who a body of evidence is relevant to, for what purposes, and what boundaries exist to generalisation beyond these conditions. 

The need to account for context does not mean one is free to choose their own adventure. On the contrary, all relevant findings must be accounted for. The quality of that evidence must also be accounted for: peer review is a minimum standard but does not replace incisive questions about how research is conducted or what, collectively, the findings can tell us. There is a sense of intellectual humility in being willing to change one’s approach in line with valid and robust evidence.

Second, evidence evolves. A recent review of well-known classroom strategies emerging from cognitive science found that some had been tested across year groups and subjects (e.g. retrieval practice), while others were tested predominantly in the middle years of schooling and in Maths or Science (e.g. interleaving). Where evidence is simply missing, and not contrary to practice, teachers and policymakers must use sound judgement to consider how relevant related evidence might be. 

Description which has grown capital letters

Given the evolution of knowledge, we should also be wary of definite characterisations of evidence that don’t appear open to nuance or change. The Science of Reading (together with other ‘Sciences of’) is an interesting example of a “description which has grown capital letters”: a linguistic phenomenon in which a field of study can, in the wrong hands, act semantically like a proper name – it becomes rigid and resistant to investigation and may no longer denote the field of inquiry to which it originally referred. Reading science, cognitive science, or psychological science are safer characterisations and offer much evidence that is useful in classrooms. 
Third, purpose matters. Evidence should not supplant philosophical discussions and sociologically informed considerations of what education is for or what it can reasonably be expected to do. Rather, evidence should support decision-making about how best to achieve specific educational goals within specific subjects and for specific learners.