This blog is inspired by another by Ruth Walker – E-coli and quality first teaching.  I’m basically trying to say the same thing.  In her brilliantly punchy post she uses an excellent analogy: when food hygiene is poor, the more vulnerable sectors of a population are most likely to suffer – the elderly and babies are more likely to get sick. But the solution doesn’t lie in addressing their needs as sub-groups; it lies in addressing the core issue: poor food hygiene.

I think this is a very important idea.  We have spent so long chasing rainbows with sub-group analysis; diving down rabbit holes; grasping at straws; playing whack-a-mole. The recent series of paradigm-busting blogs by Becky Allen have highlighted the folly of this in relation to chasing after Pupil Premium achievement through narrow interventions – largely because the probability of hitting the right students in ways that they need is pretty low.

A major part of the data delusion that has built up over recent years has been that each sub-group in a cohort should, more or less, achieve similar outcomes and that if there are GAPS – the GAPS MUST BE CLOSED.  Usually this involves identifying some plausible causal link and then trying to make the lower achieving group catch up to the higher achieving group by intervening around that link.  However, this is fraught with difficulty and data-wrongheadedness, too often driven by folk without enough mathematical confidence to apply the notion of ‘statistical confidence’ to their data.

Boys are not all the same. If boys underperform compared to girls ‘Boy-friendly’ strategies might not work for them all and may benefit the girls even more.  Ethnic groups rarely have neatly homogenous learning characteristics that fit with definable interventions and often the group size is so small as to render the data useless.  Every SEND student is different.  Higher Prior Attainment students are not a distinct group and can be half your school.  Then there is Odd Door/Even Door syndrome – the legendary cautionary tale once shared on David Didau’s blog suggesting that spurious door number ‘gaps’ might be greater than say gender or PP/non-PP gaps: you can find all kinds of patterns and differences between groups even if the cause is entirely random.

And this is the main point:  At some point ‘intervention’ really has to be simply ‘teaching’.  Given all the variables, uncertainties and unknowns, rather than chasing interventions, it is a far far better bet to focus on teaching everyone better.  To spell out Ruth Walker’s blog message, the reason why some students underachieve is likely to be a product of some deficit with the core teaching and learning process, not some special feature of it that just applies to them.  It’s just that – like babies and the elderly with e-coli – they might be more vulnerable and the symptoms will be more evident.

This might be about the overall pitch of your lessons, how well you specify the knowledge requirements of your curriculum, a behaviour management issue, the extent of your probing questioning, the amount of practice students get, how responsive you are in acting on information from your formative assessments, the quality of feedback students receive, the clarity of your explanations, your expectations in relation to recall and your habits around checking for understanding.  If you address these areas, everyone benefits and, given the time constraints, focusing on a couple of these things seems like a good bet.

The message is simple: Try to chase gaps directly and you have no real idea what will happen…it’s a false prize.   Try to teach everyone better and then your sub-groups will do better  – because they are part of everyone!