Research ethics need a new responsibility to teachers

Researchers’ lack of ethical responsibility to teachers wastes effort and harms pupils, writes Thomas Martell. So what are we waiting for to change it?

Researchers’ lack of ethical responsibility to teachers wastes effort and harms pupils, writes Thomas Martell. So what are we waiting for to change it?

14 Feb 2022, 5:00

A recent attempt by researchers to reconcile the reading wars has in fact only reignited them. Claims and counterclaims of the most evidence-informed way to teach reading abound, yet teachers are no clearer about how best to help all children learn to read. How did we get here? And what can we do to fix it?

The British Educational Research Association’s (BERA) widely used ethical guidance frames ethics as researchers’ responsibilities to stakeholders, including participants, funders and other researchers. Their core principle is to do no harm to the former, and this focus is partly due to infamous cases of abuse in human research, such as the Tuskegee Syphilis Study between the 1930s and 1970s that deliberately left hundreds of black men untreated so that researchers could study the disease, despite effective treatments being widely available for decades.

In response, research organisations set up ethical safeguards, including better oversight and training. The legacy is an ethical paradigm that focuses mainly on the people directly affected by research. But what about the far greater number of people who are indirectly affected by findings?

The failure to acknowledge this indirect impact occurs at every stage of the research process, and in his superb book, Science Fictions, Stuart Ritchie categorises four resulting types of malpractice: fraud, bias, negligence and hype. These exist on a continuum, with fraud rightly considered outrageous, but none is victimless. 

Thankfully, as Elisabeth Bik has shown, fraud is rare. (Though her work highlights that it’s often hard to spot when it does happen.) Bias, negligence and hype, however, are more tolerated. And the fact is that they have the same effect: findings are false, which misleads research users, contributes to mistrust of science and, in the case of educational research, ultimately harms pupils. As an extreme analogy, we might think of fraud as murder, and bias, negligence and hype as causing death by dangerous driving. Drink drivers do not intend to harm anyone, but their victims don’t know the difference.

Why not treat this like false advertising?

Growth mindset provides a salutary example. Professor Carol Dweck sensationalised her work, which we now know wasted the efforts of teachers and other researchers worldwide. The latter have tried and failed to replicate the large effects originally reported (including a high-quality EEF-funded trial) and the latest review hints at only a very small positive benefit. Some have also expressed concerns that Dweck’s use of statistics “finds success no matter what”.

Today, Dweck charges over $20,000 to give a lecture. Her books have sold over a million copies. She is an esteemed professor at an elite university and she won the inaugural Yidan Prize of $4 million. But what about the schools that followed Dweck’s advice, often using their pupil premium funding to chase rainbows?

Over time, attitudes to drink driving have profoundly changed. We need a similar change in attitude to bias, negligence and hype in research. This should start by giving researchers a clearer ethical responsibility towards research users, including teachers and policymakers.

BERA is currently recruiting a panel to advise on the fifth update to their ethical guidelines. This is an opportunity to re-balance our ethical thinking and get a better return on the millions we spend annually on education research.

The second step is re-shaping incentive structures. Currently, there are no consequences for polluting the research literature with findings that are unlikely to be true. In fact, it is arguably encouraged by the current ‘publish or perish’ context of academic career-building. Dweck has been gracious when challenged, but why do we research users accept being treated so badly? Why not treat this like false advertising, or stringent product recall regulations?

We should reward researchers for focusing more on relevant questions, conducting their work to the highest standards, communicating it responsibly and verifying the work of others. This would unlock the potential of research evidence for teachers and ultimately improve the life chances of children.

Who knows? It might even reconcile parties before any wars are ignited!

A new ethical responsibility – reinforced by tasty carrots, and pointy sticks – is needed.

More from this theme


Covid delays release of long-awaited phonics study

Important £1m trial study now due in 2023

John Dickens

Your thoughts

Leave a Reply

Your email address will not be published.