Lucy Rycroft-Smith
This blog post is based on the Evidence & Policy article, ‘What might constitute research misuse in knowledge brokering and how might Ofsted’s 2021 mathematics ‘research review’ be considered as an example?‘, part of the part of the Evidence & Policy Special Issue: Research (Mis)use and Mis/Disinformation in and around Education.
In a new debate article in Evidence & Policy, I argue that Ofsted, the body that inspects England’s schools, engaged in what I term research-washing when it published its 2021 Mathematics Research Review. Research-washing means “the application of a thin veneer of research credibility to a product or idea that’s being sold to schools or teachers” (Enser, 2019:23), or wearing borrowed and underserved robes of academic credibility. The result can mislead teachers, skew school inspections, and corrode public trust in evidence itself.
Why this matters beyond mathematics education
We live in an age of evidence-informed everything. Governments, regulators, and inspectorates routinely invoke “the research” to justify their decisions. This is, in principle, a good thing. Research should inform practice and policy. But this only works if those summarising research do so honestly and accurately, with nuance, and, crucially, engage with and respond to the communities they serve. Knowledge brokers (organisations and individuals who facilitate the movement of research into decisions) can wield enormous influence, and in this article, I argue that Ofsted meets the definition of a knowledge broker. Ofsted’s documents that they refer to as ‘research reviews’ are read by school leaders and teachers across England, shaping how subjects are taught and how schools are judged. Yet, there is currently no agreed standard for what good-quality and ethical knowledge brokering looks like, and no accepted definition of what it means to misuse research in this context. In this article, I propose some suggested ways in which to do this, using the Ofsted document as a case study.
A new taxonomy for research misuse
Drawing on my experience as a knowledge broker and researcher, I developed a typology of eight types of research misuse that can occur when evidence is translated for policy and practice (see Table 1). I then tested this against Ofsted (2021), using publicly available critiques from the mathematics education community as evidence.
The evidence gathered by the mathematics education community, which includes professional associations, researchers, and university groups, reveals sustained problems across at least seven of these eight categories, which you can read about in the article.
What is research-washing and why is it so effective?
Research-washing is the educational equivalent of greenwashing. Just as a company can claim environmental credentials while doing little to earn them, a document or output such as video can claim to be evidence-based while functioning as advocacy or opinion dressed in academic robes.
What makes research-washing so insidious is that it is difficult to spot without specialist knowledge. A document or output may look credible by using formal language, cite studies, and being published by a trusted authority. Teachers and school leaders, who are understandably busy, have no reason to check every footnote. And Ofsted, as England’s schools inspectorate, carries enormous institutional weight.
This is precisely how research-washing causes harm. It does not just mislead in the moment. Rather, it shapes what gets taught in classrooms, how schools are judged during inspections, and what counts as good practice in a subject, potentially for years. It also harms the community of knowledge brokers and the reputation of research in general, by eroding trust and credibility. Granted, this is in the best-case scenario, when the research-washing is revealed. It is likely that these practices are happening often, but being identified rarely. You may have noticed the striking similarities with automated processes that generate text such as large language models: text that looks plausible, draws on community protocols, but often misleads and attends only superficially to topics that require expert attention. AI is accelerating the process of research-washing by making it exponentially faster and easier to create these kinds of artefacts without the requisite checks and balances to ensure they are good quality.
What can be done?
The typology I have developed is not just a way of diagnosing past failures. It is intended as a practical tool: something that organisations publishing evidence reviews, policymakers commissioning them, and practitioners reading them could use to ask better questions and audit their own practice. For example:
- Does this document name its authors?
- Does it describe how sources were selected?
- Does it acknowledge where researchers disagree?
- Does it distinguish between what a study shows and what we might conclude from it?
- How does the organisation respond to queries or challenges from the community?
- What happens if new evidence comes to light?
These are basic standards of good quality evidence use that expert knowledge brokers have been negotiating for many decades. There is also a broader accountability question. When a knowledge broker as powerful as Ofsted produces a document that many leading researchers in the relevant field say misrepresents their work, and then refuses to correct it, there should be accountability measures that correct this and prevent it from happening repeatedly. There should be processes that stop harmful misinformation from making it into the educational ecosystem, since we know it can be particularly hard to clean up once released. These issues are made more urgent by the current availability of generative artificial intelligence tools. This may be a crisis of misinformation akin to a giant oil spill of research-washed slop, with far-reaching effects. The question is, who will clean up the mess?
IImage credit: Image by bookdragon on Pixabay
Dr Lucy Rycroft-Smith, University of Cambridge. Email: lr444@cam.ac.uk. Bluesky: @ljrycroftsmith.bsky.social.
Read the original research in Evidence & Policy:
Rycroft-Smith, L. (2026). What might constitute research misuse in knowledge brokering and how might Ofsted’s 2021 mathematics ‘research review’ be considered as an example? Evidence & Policy, DOI: 10.1332/17442648Y2026D000000098.
If you enjoyed this blog post, you may also be interested in reading:
Politics of neutrality: intermediaries and research use in civics programming
Disclaimer: The views and opinions expressed on this blog site are solely those of the original blog post authors and other contributors. These views and opinions do not necessarily represent those of the Policy Press and/or any/all contributors to this site.