total learning: shoeless learning | superclasses | rooms within rooms | write-on surfaces | toilets | schools-within-schools | sound | learner led | science spaces | phones: how young? | tiered seating | little details | flat screens | rule of 3 |. . . . ©professor stephen heppell

schools within schools: vertical groups


As the convincing evidence grows for the importance of peer support, peer to peer mentoring, of the role for mixed age support and of the need for "memberships" in advancing standards and making better learning we come, inevitably, to the question: "what is an appropriate group size?" and of course the supplementary ""and how do we achieve that if the ideal size is smaller than our school?". An answer very much in the ascendant is to move to schools within schools - in the various guises that have been seen to work, and work well. This page explores that movement.

school organisation diagram


There are lots of examples, and you may well know some from direct experience, but to give a flavour of the diversity of this approach see these few from many:

cayman net news article

The results can be rapid. In our caribbean school within 18 months there were real gains as the children explain in this pdf:

Terminology doesn't really matter; in other implementations the "schools" can be called Houses, Home Bases, Colleges and more. For me, the important bit is the clear identity - uniforms, badging, colours and more all help signify this unrequivocably to all - and the separation (Gareth and I built a fence between our new Caribbean schools, even though they shared some facilities). Hopefully, you can see that the small scale communities and large scale organisation can mesh together effectively, in different ways.


There is plenty of research evidence, summarised for example in this ERIC digest pdf. The research is unequivocal that reducing the scale of the school community within a larger school is repaid immediately - the ERIC conclusions reflect that "While research results are limited, the school-within-a-school model has the potential to contribute to a greater sense of student well-being, a sense of student community, and higher student achievement and educational attainment".

As with much in education, results are not black and white. It is never that simple - there are a lot of details to get right and some schools-within-schools efforts don't get the details right. Those details really matter.. If you've read other pages on my website you will already know the mantra that none of this is hard, but it is complex. So of course the research doesn't just say "schools within schools are fab - results skyrocket".But when the details are attended to, the results are quite remarkable. As the Gates Foundation observed: "The evidence is clear that smaller impersonal schools are no more effective than larger impersonal schools". Turning a large factory into a smaller factory misses the point! and although Gate's work wobbled on getting the details right early on (a school specialising in firefighters for example!), the results in the end were signicant gains. See for example New Study Gives Small Schools Initiative a Thumbs Up from 2012.

Here are a couple more useful links: a BBC report on the Calouste Gulbenkian Foundation research on Schools within Schools, and an older summary of the research evidence in the US - in the US evidence you will note the more recent research evidences the best academic gains. This is presumably because the details are better understood and wider circulated. The peer to peer model of teacher CPD that is common in the 3rd millennium may be better at transmitting these key details than the top down model we had before, perhaps. Or it may be that developments like superclasses reinforce and "multiply" the collegiality and sense of community.

Even when the important details are not all implemeted however, schools within schools always deliver on a better experience for children. Even back in 1995 Greenleaf's research noted that smaller units of learning communities "increased their social commitment to one another and to their teachers, thereby increasing their personal investments in school".

There is a lot of good research to suggest that there is a cognitive limit to the number of people with whom one can maintain stable social relationships.

Dunbar's number
British anthropologist Robin Dunbar suggested that the optimum number to maintain a stable, cohesive group (not just people that know each other, but folk who share a stable inter-personal relationship) lies between 100 and 230. This is a quite precise number, with an "ideal" number often quoted as 150. It is precise because Dunbar theorised that there was a cognitive limit - this was all the brain has eveolved to be able to deal with.

Dunbar said: "this limit is a direct function of relative neocortex size, and that this in turn limits group size ... the limit imposed by neocortical processing capacity is simply on the number of individuals with whom a stable inter-personal relationship can be maintained."

You could have a lot more people in your life - other aquaintances, people you have lost touch with, but the limit on this larger number would probably depends on long-term memory size. These are not the people you have a relationship with, and this relationship is what is important to communities of purpose like schools. You need stable and inter-personal relationships to learn and the limit is imposed by your brain: 230 tops.

Like much else in cognitive science, it has been largely proposed since the 90s and our use of fMRI, or functional magnetic resonance imaging. Others have suggested that Dunbar's number applies to on-line communities too. Some have suggested a higher number (Bernard and Kilworth settle, after much work, on a mean of 290). But it is interesting to note that groups facing tough circumstances - for example subsistence villages, or military groupings in history, seem on average to hit something pretty close to the 150-member limit.

You can see Robin giving a short and entertaining talk at the RSA.


Of course, the question is: how to achieve these importantly small numbers, but still allow what Dundar and others call the "social grooming" that binds a community together, and to allow young to learn from old effeciently, without coercion or disengement? The answer does seem to be the Schools within Schools movement. Splitting a large school into discreet vertical groups allows that continuity of community together with a clear sense of collegiality, of mutuality, of belonging. And everyone knows the whole, small, community members well. I asked a student in a school we had divided into Schools within Schools, why she thought it worked so well. memorably she replied that:

"it's not just that they know you dog has died, it's that they already knew you had a dog, and what i's name was"

I thought she had grasped it very well. Several cities, including New York City, Philadelphia, and Chicago, have school within school projects running - in Europe there are a rapidly growing number too.


In the literature, the most critical factor for success is a commitment to implementing the program fully, allowing for proper separation of the subschool and the creation of a clear separate identity (McCabe & Oxley, 1989; McMullan, Sipe, & Wolfe, 1994; Raywid, 1996) - probably Dunbar's work in part explains this, but also a huge gain is the ability for each component school to be able to take a different focus (boys' performance, parental engagement, literacy gains, etc) and focus on that so that the overal institution can rapid-prototype four or five ideas in parallel and learn more quickly what works and what is effective. Either way my own experence is also that a very clear separation, with strong markers for membership, works significantly better. The stronger the separate identities, the better the gains. Half baked is not enough.

Architecturally, the schools within schools need a focus for their identity: a quad, a shared eating area, design and colour... and the "members" spend most of their learning time together, core subjects especially.

Organisationally, things vary. BUT, the heads of the schools within schools are the key personnel, the senior team of the overall school. They set the tone for 'their' school / house / family. A little affectionate rivaly between them is helpful too!


page last updated by Stephen Heppell - Saturday, February 14, 2015 5:44 PM