Schools are not nearly as uncomfortable with sorting students as they like to pretend. We talk about inclusion, belonging, mixed-ability learning, and meeting every student where they are, which are all noble enough ideas when they are taken seriously. At the same time, we run selective schools, opportunity classes, gifted pathways, accelerated streams, extension groups, intervention groups, support classes, withdrawal programs, and every other structure that allows us to say we believe in inclusion while quietly moving certain students somewhere else. This is not necessarily hypocrisy. Schools are complicated places, and students do not all need the same thing at the same time. The problem is not that schools group students. The problem is that schools often group students without being honest about the philosophy behind the grouping.
That is where streaming becomes interesting, because the usual debate is far too predictable. One side argues that streaming allows teachers to target instruction more precisely, reduce the range in the classroom, and provide students with work better suited to their current level. The other side argues that streaming locks students into hierarchies, lowers expectations, damages confidence, and reproduces advantage. Both sides have a point, which is inconvenient for anyone hoping for a simple answer. A teacher who has stood in front of thirty students spread across five years of readiness can understand the attraction of streaming. The accomplished student is bored. The struggling student is drowning. The teacher is somewhere in the middle, trying to preserve order, dignity, and possibly a small part of their soul.
So, yes, streaming has a logic. If students differ significantly in current achievement, prior knowledge, pace, confidence, and skill, then grouping them by readiness might make teaching more focused. That is not a ridiculous idea. In fact, it may be quite reasonable in some contexts. But if this is the belief, then schools should have the courage to follow the logic properly. Stream seriously. Stream coherently. Stream with conviction. Do not just pull out the obvious top group and the obvious bottom group, call the leftover group “mixed ability”, and then pretend a serious pedagogical decision has been made.
This is my issue with much of what passes for streaming in schools. It is not streaming as a philosophy; it is selective extraction. The top students are removed into an accelerated class, the most visibly struggling students are placed into a supported class, and everyone else remains in the broad middle, which is still diverse, still uneven, and still full of students with different levels of readiness. The middle is treated as if it is now coherent because the extremes have been lifted out. But the middle is not coherent. It is simply what remains after the easiest students to identify have been moved elsewhere.
If schools really believe that students learn better when grouped by current level, then why stop with Math and English? Why not stream everything? Stream Science, History, Art, Music, PDHPE/PE/Gym, and anything else on the timetable. If the principle is sound, it should apply beyond the subjects most easily tested or most anxiously watched by parents. Should the accomplished pianist sit in the same music class as the student who has never touched an instrument? Should the child who has been taking Saturday morning art classes since they were three be assessed alongside my stick figure drawings, which, frankly, even the stick figures would find insulting? Should the school’s star swimmer be in the same group as the student who still holds their nose, negotiates with gravity, and enters the pool as if signing a legal waiver?
These examples sound comic, but they expose the inconsistency. Schools often stream where achievement is measurable, public, and tied to status. Math and English are the obvious candidates because they are treated as core subjects, closely linked to school reputation, parent anxiety, standardised testing, and the machinery of academic sorting that begins early and pretends to be neutral. Science may be added if the school is ambitious or the timetable allows. Humanities, the arts, and PE are less often treated this way, even though skill differences in those subjects can be just as real and sometimes far more visible. The student who has played violin since age four and the student meeting a treble clef for the first time are not merely at different levels; pedagogically, they are barely in the same time zone. The student who trains six days a week and the student who considers running a personal attack are not at the same level. Difference does not politely confine itself to NAPLAN adjacent subjects.
Of course, streaming everything would be a logistical nightmare, which is one reason schools rarely do it. Timetables would become harder to build, parent anxiety would intensify, and movement between levels would need to be frequent enough to be credible without becoming a public ranking ceremony. Teacher allocation would also become political, because a school that believes in streaming would have to confront an uncomfortable question: are the students with the greatest learning needs getting the strongest teaching, or just the smallest room and the newest staff member? Every subject would need a defensible placement process, and every placement process would generate emails, appeals, wounded pride, and at least one parent explaining that their child’s brilliance has once again been misunderstood by the institution. However, if the philosophical commitment to streaming were strong enough, schools would make these systems work and defend them unapologetically. The fact that most do not is understandable, but it also reveals the problem. Many schools want the advantages of streaming without accepting the full cost of the idea.
That is why this is useful as a thought experiment. I am not convinced that schools should stream everything. I am not even convinced that streaming is the right default. But if we believe in streaming, then we should at least understand what the belief demands. A serious streaming model would require movement. It would require regular review. It would require humility about what any test or teacher judgment can capture. It would require the school to guard against lower streams becoming zones of lower expectations. It would require the strongest teachers to work with the students who need the most expert instruction, not just the students who make the school look good on presentation night. Most importantly, it would require schools to admit that streaming is not a timetable trick. It is a philosophical position with consequences.
The same is true on the other side. If a school believes in heterogeneous inclusion, then it should lean into that belief properly. It should stop quietly extracting students whenever the model becomes difficult. It should resource differentiation, provide planning time, support teachers with specialist expertise, and stop pretending that mixed-ability teaching works because the phrase sounds morally attractive. Inclusion is not the cheap option unless we are doing it badly. A genuinely inclusive model requires serious professional trust, thoughtful curriculum design, smaller classes where possible, and the conditions that allow teachers to know their students well enough to respond to them as individuals.
This is where schools often fail. They want the moral language of inclusion and the status benefits of selection. They want the flexibility of mixed-ability classes and the parent-pleasing optics of top streams. They want to say every child belongs, while also running structures that signal, very clearly, which children belong in the high-prestige rooms. They want competition without the brutality of open ranking, and inclusion without the cost of making inclusion work. So, they settle for a compromise that offends the fewest people in the short term but satisfies neither philosophy in the long term.
The regular class is where this compromise becomes most obvious. It is often described as mainstream, but in many cases, it is really the residual class, the broad middle left behind after selective extraction has occurred. It contains the quiet struggler, the lazy high achiever, the student with gaps from primary school, the student who reads like an adult but writes like they are texting while riding a bike, the capable student who has discovered that doing the minimum produces roughly the same consequences as trying, the anxious perfectionist, the slow worker, the charismatic avoider, and the student whose entire academic strategy is to wait for the bell. Calling that group “mixed ability” is technically accurate but not especially helpful. It describes the variety, but not the problem.
Nor are the streamed classes themselves as coherent as schools like to imagine. The accelerated class is not a collection of identical miniature scholars waiting politely for a university reading list. It includes students who test well, students with strong prior knowledge, students with substantial support at home, anxious perfectionists, quick but shallow thinkers, and students who are coasting because school has not yet asked very much of them. The supported class is just as varied. It may include students with learning disabilities, students with gaps from disrupted schooling, students who lack confidence, students who struggle with language, students who are bored, and students who are more capable than their results suggest. Streaming narrows the difference in some ways, but it does not eliminate the difference; it rearranges it.
That point matters because it connects directly to the previous post about differentiation. Even in a fully stratified school, streaming does not release teachers from the obligation to know students as individuals. It might narrow the instructional range in each room, but it would not make the students interchangeable. An advanced English class will still include the fluent reader whose writing lacks precision, control, or flair. The lower class will still include students learning English as an additional language, alongside students who are perfectly capable but have learned that doing very little is usually enough to survive. Streaming can organise differences, and sometimes that may help. What it cannot do is abolish it. The group may become more coherent, but it never becomes a singularity.
This is also where the research caution matters. Ability grouping has never received the clean endorsement its strongest defenders sometimes imply. Effects are contested and often modest, and the risks around expectation are real. Once a student is placed in a group, the group can begin to speak louder than the student. Teachers may expect different things. Students may internalise different identities. Parents may read placement as destiny. Schools may insist movement is possible, while everyone involved quietly understands that movement is rare, awkward, or socially loaded. A stream that begins as a flexible instructional grouping can easily harden into a status system.
That does not mean all grouping is wrong. It means grouping should be treated as a serious educational decision rather than a convenient administrative habit. Sometimes students need a different pace, more explicit instruction, greater challenge, additional practice, or a temporary group formed around a specific need. There is nothing inherently immoral about that. What is questionable is the conversion of temporary grouping into permanent identity, or the use of partial streaming to create the appearance of precision while leaving the deeper problem untouched.
The deeper problem is that schools often refuse to choose. If they believe in streaming, they should stream with conviction and accept the costs: transparency, movement, stigma management, parent discomfort, teacher allocation, and the constant risk that hierarchy will become destiny. If they believe in heterogeneous inclusion, they should also act with conviction and accept those costs: planning time, workload relief, specialist support, serious differentiation, and trust in teachers’ professional judgement. What does not work is pretending we can have both without paying for either.
Streaming without conviction is just selective extraction. It removes the students easiest to identify, rearranges the problem of difference, and leaves schools congratulating themselves for having made a decision they have not really made. If we are going to sort students, then we should be honest about what we are doing and why. If we are not, then we should stop building half-streamed systems that neither honour inclusion nor pursue streaming seriously. Either way, the central obligation remains the same: the group is not the student. A stream, set, band, class, pathway, or program may tell us something useful, but it never tells us enough. The work of teaching still begins when we look past the category and notice the person sitting in front of us.
Maximising Student Potential 