Guest Post: Cherry picking or broad vistas: Using research evidence to improve outcomes for vulnerable learners

Schools can improve outcomes for vulnerable pupils by becoming more research informed, argues Marc Rowland, Head of Rosendale Research School…

The main thing I have learned from working with schools on improving outcomes for disadvantaged and vulnerable learners is that there is no magic elixir. There is no sledgehammer that can break the shackles of underachievement in vulnerable groups. Rather, schools need to create a complex, delicate tapestry of strong values, high expectations, inclusion, knowledge of pupil needs, community needs and staff needs. They need high quality teaching which has the capacity and expertise to deploy resources where they are needed most. They need a language rich, culturally rich curriculum. Where intervention is needed, it should be high quality, structured, time limited and targeted, with accountability remaining with the teacher. They need robust quality assurance, sharp impact evaluation and a balance of accountability to external organisations with accountability to pupils and families.

Wherever possible, actions we take to weave together a strategy to overcome educational disadvantage should always be underpinned by evidence.

Engagement with evidence has come a long way in the last 10 years. In 2008, Estelle Morris, Jonathan Sharples and a small band of warriors at the Coalition for Evidence-Based Education were rather lonely voices. Much has changed. The Education Endowment Foundation has transformed the educational lexicon, raising the profile research evidence like the Himalaya from the Sea of Tethys. But whilst an awareness of the EEF toolkit is necessary, it is not sufficient. Firstly, teachers and leaders must avoid the temptation to be overly meta. Because that risks oversimplification, missing delicate nuance, subtlety, context and contradictions in research evidence. Secondly, however tempting, we must avoid cherry picking the headlines from evidence to support decisions we have already made, rather than use evidence to inform decision making. Cherry picking evidence because it supports our own preconceived biases simply provides a veneer of respectability for decisions. But the impact will be limited.  

We must retain broad vistas about education research. We need to be looking at evidence beyond what is immediately and easily accessible, being open to new thinking and new findings, however challenging and uncomfortable. The University of Durham and UCL London are a big rock candy mountain of useful, information, challenging and reassuringly disconcerting sources of evidence. Professor Rob Coe’s ‘Improving Education’ remains a masterpiece of challenging orthodoxies. ‘Poor Proxies for Learning’, ‘Mistaking School Improvemnet’ and ‘What types of CPD impacts on learners’ should be known and understood of everyone working in education. Awareness of the work of  Waldfogel and Washbrook in 2010 (Sutton Trust) highlights the need to tackle the causes of educational disadvantage, rather than just the symptoms if we are to make a breakthrough:

Vocabulary at aged five:

  • There is a 27% gap between the lowest income quintile and the highest.
  • The lowest quintile have 16% more likely to have conduct problems compared to the highest quintile.
  • The lowest quintile are 15% more likely to have hyperactivity problems compared to the highest quintile

Waldfogel and Washbrook, 2010

Zero tolerance strategies for behaviour are boats against the current if we don’t tackle the vocabulary gap.

And as pupils get older, access to research evidence shows us we can still treat the causes…

The Millennium Cohort Study from ICL Institute of Education shows:

“Analysing the scores of nearly 11,000 14-year-olds in a word exercise, the researchers found that teenagers who read for pleasure every day understood 26 per cent more words than those who never read at all in their spare time. And teenagers from book-loving homes knew 42 per cent more words than their peers who had grown up with few books.

Even taking into account other factors, like parents’ qualifications and profession, and cognitive tests taken by the teenagers when they were aged 5, teenagers who read for pleasure still got 12 per cent more words right, while those from book-rich homes scored 9 per cent more.” 


Engaging with research evidence shows us that building children’s vocabulary unlocks the gates of learning.

But we must go further afield than that. An old University lecturer of mine, Dr Brian Dicks, used to tell me ‘there’s a lifetime of travel in Europe, you don’t need to go further than that’. Dr Dicks is almost certainly right that you could spend a lifetime travelling Europe, but as someone who’d only been abroad to France twice at the age of 20, the experiences I’d have missed out on if I’d listened to that reassuring view that sat comfortably with my own limited horizons is almost too much to think about. So we need to look beyond what we know. One of my favourite pieces of research in the last year comes from the University of Missouri:

Students More Likely to Succeed If Teachers Have Positive Perceptions of Parents

 Published: February 21, 2017.

Released by University of Missouri-Columbia 

“It’s clear from years of research that teacher perceptions, even perceptions of which they are not aware, can greatly impact student success,” Herman said. “If a teacher has a good relationship with a student’s parents or perceives that those parents are positively engaged in their child’s education, that teacher may be more likely to give extra attention or go the extra mile for that student. If the same teacher perceives another child’s parents to be uninvolved or to have a negative influence on the child’s education, it likely will affect how the teacher interacts with both the child and the parent.”

Whilst we can’t always avoid bias (and perhaps, neither would we want to), by being aware of the impact of bias, we can adjust our behaviours to negate any negative effects. Getting little things wrong, over the lifetime of a child’s education, matter a great deal, particularly for those from disadvantaged backgrounds.

And finally… However robust the research evidence in front of you, it is worth remembering that the quality of implementation and school / class context is fundamental to the effectiveness of any intervention. Avoid the Batsian mimicry approach. What might look like an effective intervention in a nearby school, with a reassuringly expensive lever arch file might not be all it seems. If it is not targeted at the specific learning needs of your pupils (and by this, I do not mean attainment levels), it is unlikely to succeed.


You can visit Rosendale School at our forthcoming event with Rob Webster, Kate Atkins and David Bartram OBE:


What possible evidence does Ofsted chief have to suggest religious extremists are ‘perverting’ education?
Duchess of Cambridge launches new campaign to raise self-esteem of kids
Categories: Guest Post.


  1. Interesting article. The EEF have produced enormous quantities of data and have been generously funded by the tax-payer to do so. Based on my own experience, the reports do not necessarily tell schools exactly what happened during the trials. For example, if the pupils who participated in the project were tested 8 weeks after the project ended, then the EEF report should acknowledge this. A delay over the summer holiday impacts the quality of the data. Some EEF funded projects require pupils take part for an hour per week and some for ten minutes per week. If they both produce the same progress for the same money, which intervention has more impact? Is that something that should be addressed – I think so – but perhaps I’m biased? Rhythm for Reading is not an Arts Participation programme as EEF tell you. It is a reading programme.

Let us know what you think...