My IB Why This Matters
"I gave them a completely open source-evaluation task in week one because I wanted them to inquire, and it was chaos — half of them just quoted the source instead of evaluating it." That's the failure of unscaffolded inquiry: dropped in the deep end, students without the skill flail or default to the simplest possible response. But the opposite failure is quieter and just as damaging: "I've been giving them the same sentence starters and guided worksheet since September, and now it's May and they still can't evaluate a source without them." Both problems come from the same missing decision — nobody planned when the support would stop.
Understand the Idea
Scaffolding is temporary support that helps a student do something they can't yet do independently — a sentence starter, a partially completed example, a checklist, a worked model. The word matters: scaffolding, on a building, comes down once the structure can stand on its own. Support that never gets removed isn't scaffolding anymore, it's a permanent crutch, and the whole point of inquiry-based teaching is that students eventually run the inquiry themselves.
The planning tool for doing this deliberately is the gradual release of responsibility — a sequence that moves, in stages, from the teacher doing the thinking out loud, to the teacher and student doing it together, to the student doing it with light support, to the student doing it alone. It's sometimes summarised as "I do, we do, you do." Each stage should be planned in advance, not improvised in the moment based on how tired you are that day.
The first stage typically involves modelling: you perform the skill yourself, out loud, in front of students, narrating your thinking as you go, often alongside a worked example — a completed sample showing exactly what a strong finished product looks like, annotated to show why each part is strong. Students don't yet do the skill; they watch you do it and see the standard.
The middle stage is guided practice: students attempt the skill with support still present — a checklist, a partially worked example they complete the second half of, a peer to check against, or you circulating and prompting rather than doing it for them.
The final stage is independent practice: the scaffold is removed, and students do the skill unaided, ideally on a task similar in structure to what they'll eventually face in the summative assessment.
This entire sequence should be planned around a specific ATL skill — one of the transferable skills, like evaluating sources, organising an argument, or managing time, that you identified when planning your unit in Chapter 03. Scaffolding without a named target skill turns into general hand-holding with no clear endpoint. Naming the skill is what lets you decide, in advance, what "independent" will actually look like and by when.
See It in Practice
Take a three-week unit built around the ATL skill of source evaluation — judging whether a historical or scientific source is reliable, biased, or useful for a given purpose.
Week 1 — modelled. The teacher projects a single source (a nineteenth-century factory owner's diary entry) and thinks aloud, live, in front of the class: "First I'm checking who wrote this and why — a factory owner defending long working hours is not a neutral witness, so I'm already suspicious before I've read a word of content. Now I'm checking the date against the events it describes — is this written at the time, or years later from memory? That changes how much I trust the detail." Students copy nothing yet except the structure of the thinking, not the content — three questions the teacher used, written on the board as a template: Who wrote it and why? When, relative to events? What's missing?
Week 2 — partial/guided. Students receive a new source and the same three-question template, but this time they apply it themselves, in pairs, with the teacher circulating and prompting rather than answering: "You said this source seems reliable — what in the source itself makes you say that, versus what are you assuming?" Some pairs get a partially completed example alongside their own source, showing one question answered in full, so they can see the depth expected before attempting the remaining two themselves.
Week 3 — independent. Students receive a new, unfamiliar source with no template visible, and are asked to evaluate it under conditions close to the summative task — individually, within a set time, using only their own notes from weeks one and two if they choose to. The teacher does not circulate to prompt during this session; this is a checkpoint for how much has actually transferred, not another guided round.
Weak version of this same unit gives students the three-question template in week one and never removes it — by week three, "guided" and "independent" look identical, still worksheet-supported, and the summative assessment (which offers no template) becomes the first time students meet the task unsupported. That's not scaffolding faded, that's scaffolding left up until moving day, which is when students discover the wall behind it was never actually built.
Strong version, shown above, has a visible, planned difference between week one and week three: modelling with full teacher narration, then partial support with prompting, then true independence, timed to land before the summative, not during it.
Apply It
Write two short paragraphs, no more than four or five sentences each:
What week one support looks like — describe the modelling moment specifically: what you say out loud, what worked example you show, what template or structure you hand over.
What week three independence looks like — describe the task students face with the scaffold removed: what's different from week one, what conditions (timed, unfamiliar material, no template) make it a genuine test of transfer rather than a repeat of the guided version.
Avoid This
Symptom: the scaffold never leaves — students hit the summative task and see an unsupported format for the first time. Fix: build a stated fade-out point into the planner before the unit begins, and treat removing the support as a planned event, not a hopeful assumption.
Symptom: the scaffold is removed all at once, with no guided middle stage. Students go from fully modelled to fully independent in one lesson, and the gap is too wide to cross. Fix: keep the "we do" middle stage — partial support, prompting rather than answers — even if it costs a lesson you'd rather spend on content.
Symptom: "modelling" is really just telling students the answer, not showing the thinking. If week one is "here's what a good source evaluation says" rather than "here's how I work out whether a source is good," students copy the conclusion without learning the process. Fix: narrate your thinking, including the false starts and reconsiderations, not just the polished result.
Check Yourself
For your target ATL skill this unit, can you name the exact lesson where support disappears, and describe what students face at that point without a scaffold? If your answer is "they'll be ready by then," that's a hope, not a plan — go back and name the lesson.
Continue Learning
→ Next: Article 20 pulls the whole chapter together into one repeatable frame for planning a single lesson, not just the unit around it.