In today’s educational climate, the phrase ‘what works’ has become something of a mantra. It refers to the use of rigorous, often quantitative, research methods – particularly randomised controlled trials and meta-analyses – to identify interventions that have been empirically shown to produce measurable improvements in outcomes (Saunders & Whelan, 2022).

This approach to evidence-based practice in education is increasingly held up as the gold standard for making informed decisions. Proponents argue that teaching should be guided by interventions proven effective through systematic reviews, randomised controlled trials and large-scale meta-analyses (Slavin, 2002; Nevill, 2016).

At first glance, this movement seems like a welcome corrective to the prevalence of ‘edumyths’ – widespread beliefs in education that lack empirical support, such as the idea that right-brain and left-brain learners require different pedagogical strategies (Dekker et al., 2012; Bennett, 2013). Yet while we should rightly challenge pseudo-scientific claims, the uncritical embrace of the ‘what works’ approach risks falling into the very trap it seeks to escape: the creation of a new myth.

As Theodor Adorno and Max Horkheimer (1944/2002) warn in Dialectic of Enlightenment, Enlightenment rationality, when taken to its logical extreme, can collapse into its opposite. In the effort to demystify and control the world, reason becomes instrumental – concerned more with utility and efficiency than with understanding or human flourishing. Paradoxically, the very drive to eliminate myth through reason ends up producing new myths in the form of rigid systems, ideologies, and unquestioned ‘truths.’

The Enlightenment and its educational echoes

Adorno and Horkheimer’s critique of Enlightenment thought resonates powerfully – I think – with current trends in educational research and policy. Just as Enlightenment reason sought to categorise, measure and control nature, evidence-based education often attempts to systematise and quantify learning in ways that mirror this instrumental logic.

Teaching strategies are ranked according to effect sizes (Hattie, 2009; Education Endowment Foundation, n.d.), classrooms are evaluated against key performance indicators and interventions are scaled up or discarded based on statistical significance. What gets lost in this technocratic approach is the recognition that education is a deeply contextual, relational and value-laden practice.

Adorno and Horkheimer would argue that such a narrowing of vision amounts to a kind of regression. It could be argued, then, that instead of liberating teachers and learners, the promise of rationality becomes a form of control – a myth that reduces complex human phenomena to input-output models.

The seduction of certainty

One of the appealing aspects of the ‘what works’ discourse is its promise of certainty. Faced with the challenges of classroom diversity, curriculum overload and accountability pressures, teachers are understandably drawn to clear, evidence-backed answers. But as Biesta (2007, 2010) points out, the question of ‘what works’ is incomplete without also asking: for what purpose? For whom? In what context?

What works in one context may not transfer meaningfully to another. The effectiveness of a strategy like direct instruction, for example, depends not just on the method itself but on the teacher’s pedagogical intent, student needs, school culture and broader societal values. To assume otherwise is to fall into what Adorno and Horkheimer might describe as a new form of positivism – a belief that the world can be wholly understood and governed through data. Therefore, a good piece of research will always be aware of – and acknowledge – its limitations, which might not be evident when championed by more popular advocates of strategies like direct instruction (see Wyse et al., 2018, for a wider discussion on this issue).

This faith in technical solutions can obscure the interpretive, moral and political dimensions of teaching. The teacher becomes less a reflective practitioner and more a technician implementing prescribed methods – what Biesta (2010) terms the ‘learnification’ of education, where the richness of education is reduced to the measurable acquisition of knowledge and skills.

Other researchers have pointed this out too, with Ball (2021) suggesting that teachers are becoming mere technicians and a recent article by Pierlejewski, Murtagh and Humphreys (2024) suggesting – in a similar vein to Biesta – that ‘trainification’ and ‘pupilification’ is reducing the agentic complexities of teachers and students alike to mere configurations of simplified performance indicators.

Edumyths and the myth of demythologization

Critiquing edumyths is undoubtedly a valuable and necessary task. However, the fervour with which some educational commentators pursue this mission can at times verge on dogmatism. Blogs, articles and teacher training resources increasingly feature lists of ‘debunked’ practices – yet the rigour of such claims is not always assured. One set of materials I encountered, for example, misattributed theories to the wrong scholars and confidently asserted that a particular pedagogical approach had been ‘completely debunked,’ despite there being no scholarly consensus to that effect.

A further concern is that many of the most prominent voices in the edumyth discourse lack formal research training, or work within narrowly defined disciplinary paradigms that prioritise certain forms of evidence – typically quantitative and experimental – while marginalising others, such as qualitative, interpretive or critical methodologies. While often well-intentioned, this style of myth-busting risks reinforcing a reductive binary: that educational practices are either scientifically validated or entirely without merit, leaving little room for complexity, context or professional judgement.

The problem is not that myths are challenged, but that the process of myth-busting itself becomes mythologised. The figure of the ‘rational educator’ emerges as someone who slavishly follows the data, avoids all ‘irrational’ practices and distrusts intuition or tradition unless scientifically verified. As Adorno (2005) warns, such fetishisation of rationality can itself become irrational, leading to what he calls the ‘jargon of authenticity’ – a hollow performance of reason devoid of substance.

Moreover, what counts as a ‘myth’ is often defined narrowly by what can be tested using specific methodological tools. Practices that are difficult to quantify – such as the cultivation of moral judgement, the development of student-teacher relationships or the nurturing of democratic dispositions – are marginalised because they do not fit the preferred evidence hierarchies (Biesta, 2015).

Reclaiming judgement and complexity

What is needed, then, is not a rejection of evidence but a broader conception of what constitutes valid educational knowledge. As Gert Biesta (2020), Stephan Ball (2021) and others have argued, education is a normative, not just a technical endeavour. It involves making judgements about what is educationally desirable, not just what is effective. These judgements cannot be outsourced to algorithms or meta-analyses; they must be made by educators situated in real-world contexts.

A more dialectical approach to evidence-based practice would embrace contradiction, context and contingency. It would recognise that good teaching often involves navigating tensions between competing aims – freedom and structure, tradition and innovation, care and rigour. And it would treat evidence not as a mandate, but as a resource for deliberation and reflection.

Conclusion: Toward a critical pedagogy of evidence

Adorno and Horkheimer’s Dialectic of Enlightenment reminds us that the quest for mastery through reason can lead to new forms of domination if left unchecked. In the realm of education, the ‘what works’ approach risks becoming a myth of its own – a myth that suggests all important questions can be settled by data, that complexity can be tamed and that education can be engineered like a machine.

This is not a call to return to intuition or tradition alone, but an invitation to treat evidence critically, contextually and ethically. Instead of asking only what works, we should also ask: What matters? What do we value? And who gets to decide?

In reclaiming these questions, we move toward a more humane, reflective and truly enlightened vision of education – one that honours both the insights of research and the irreducible complexity of human experience.

Who were Adorno and Horkheimer?

Theodor Adorno and Max Horkheimer were German philosophers and social theorists best known for their work as leading figures of the Frankfurt School – a group of thinkers who developed critical theory in the mid-20th century. Writing in the aftermath of fascism and amid the rise of mass consumer society, they explored how Enlightenment reason, originally intended to liberate humanity, could paradoxically become a tool of domination. In their influential work Dialectic of Enlightenment (1944), they argued that rationality, when reduced to efficiency and control, risks producing new forms of myth, conformity and social control under the guise of progress. Their insights remain vital for critiquing modern institutions, including education.

References

Adorno, T. W. (2005). Critical models: Interventions and catchwords. Columbia University Press.

Adorno, T. W., & Horkheimer, M. (2002). Dialectic of Enlightenment: Philosophical fragments (E. Jephcott, Trans.). Stanford University Press. (Original work published 1944)

Ball, S. J. (2021). The education debate (4th ed.). Policy Press.​

Bennett, T. (2013). Teacher proof: Why research in education doesn’t always mean what it claims, and what you can do about it. Routledge.​

Biesta, G. (2007). Why “what works” won’t work: Evidence‐based practice and the democratic deficit in educational research. Educational Theory, 57(1), 1–22.

Biesta, G. (2010). Good education in an age of measurement: Ethics, politics, democracy. Routledge.

Biesta, G. (2015). On the two cultures of educational research, and how we might move ahead: Reconsidering the ontology, axiology and praxeology of education. European Educational Research Journal, 14(1), 11–22.

Biesta, G. (2020). Educational research: An unorthodox introduction. Bloomsbury Academic.

Dekker, S., Lee, N. C., Howard-Jones, P., & Jolles, J. (2012). Neuromyths in education: Prevalence and predictors of misconceptions among teachers. Frontiers in Psychology, 3, 429.

Education Endowment Foundation. (n.d.). Teaching and Learning Toolkit. https://educationendowmentfoundation.org.uk/education-evidence/teaching-learning-toolkit

Hattie, J. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. Routledge.

Nevil, C. (2016). Do EEF trials meet the new ‘gold standard’? Education Endowment Foundation. https://educationendowmentfoundation.org.uk/news/do-eef-trials-meet-the-new-gold-standard

Pierlejewski, M., Murtagh, L., & Humphreys, H. (2024). The simple view of teaching: authorised pedagogies, curriculum and the neoliberal learner in preservice teacher education. Asia-Pacific Journal of Teacher Education, 53(1), 65–79.

Sanders, M., & Whelan, E. (2022). What Works, faster: Towards a “rapid method”. King’s College London, The Policy Institute.

Slavin, R. E. (2002). Evidence-based education policies: Transforming educational practice and research. Educational Researcher, 31(7), 15–21.

Wyse, D., Brown, C., Oliver, S. & Poblete, X. (2018). The BERA close-to-practice research
project: Research report. London: British Educational Research Association. Retrieved from:

Picture credit: Wikicommons (Used under a Creative Commons Licence)