The hard part of using a question bank isn’t pressing “assign.” It’s deciding which questions, at what level, for whom — before you press anything. Two teachers can have the same bank and the same class, and one sets a task that nudges grades while the other sets thirty students the same six questions and gets the same disappointing spread back. The difference is selection.

This guide is about that decision. Not the logistics of distribution (the click-by-click of getting a task to your class is covered in the assign-a-quiz walkthrough), and not homework strategy in general (that’s the smarter-homework flagship). This is purely the pedagogical craft of choosing past paper questions by topic and difficulty — the filters you actually use, and how to reason with them when you assign past paper questions to students who are nowhere near identical.

The two dials that do the work

Every good question-selection decision runs through two dials: topic (what knowledge or skill the question tests) and difficulty (how demanding the question is for a student who has that knowledge). On a well-tagged bank, topic usually splits further — board, syllabus topic, sub-topic, and often command word — and difficulty is a banding the exam board’s own grade boundaries imply: roughly foundation/grade-4-ish, secure/grade-6-ish, and stretch/grade-8-plus.

Most teachers under-use both dials. They filter to a topic and take whatever order the questions come in, which means a struggling student hits a 6-mark “evaluate” on question two and shuts down, while a strong student coasts through recall they mastered weeks ago. Turning both dials deliberately — and turning difficulty independently for different students — is the whole skill.

Match the question to the moment in the teaching cycle

Before you even think about individual students, the difficulty you choose should track where you are in teaching the topic. The same sub-topic wants very different questions on Monday than it does the Friday before the mock.

• Just introduced (recall check): low difficulty, narrow command words — “state,” “name,” “identify,” short MCQ. You’re checking the basics landed, not testing application. Three or four quick questions is plenty.
• Mid-topic (consolidation): medium difficulty, “describe” and “explain.” Now you want to see whether they can link ideas, not just retrieve them.
• End of topic (application & exam readiness): high difficulty, “calculate,” “analyse,” “evaluate,” multi-step structured questions. This is where you find out who’ll actually score the marks under exam conditions.

If you assign past paper questions to students at the wrong point in this cycle — a “discuss the limitations” question the lesson after you introduced the concept — you learn nothing useful, because failure could mean anything. Matching difficulty to the moment keeps the signal clean.

Build a difficulty ramp inside a single task

Even one task to one student is better when it climbs. A ramp — easy questions first, hardest last — does two things: it gives every student a confident start so they engage, and it produces a diagnostic gradient. When you get the marked task back and see exactly where a student fell off the ramp, you know their ceiling on that sub-topic, not just a single pass/fail.

A clean six-question ramp on one sub-topic might be: two recall (low), two “explain” (medium), two applied or extended (high). Pull them by setting the topic filter to the sub-topic, then taking two from each difficulty band. It takes about as long as scrolling a textbook, and the result tells you far more than six questions of uniform difficulty ever would. (This is the natural follow-on to building topical tests one sub-topic at a time — same precision, applied to assigned practice.)

Target the gap, not the topic

The most wasteful assignment is the one aimed at a whole topic when the class only struggles with one corner of it. If your last check showed the class handles photosynthesis fine but collapses on the limiting-factors graphs, don’t reassign “photosynthesis.” Filter to the sub-topic limiting factors, pick the command words that exposed the weakness (“explain the effect of,” “describe the trend”), and set six questions on exactly that. You’ve turned a vague “revise the topic” into a precise drill on the thing actually costing marks.

This is where sub-topic and command-word filtering earn their keep. “They’re bad at chemistry” is unactionable. “They drop marks on ‘explain’ questions about rates of reaction” is a 90-second assignment that hits the real gap. Selecting past paper questions by topic and difficulty at that resolution is the difference between practice that feels like revision and practice that moves a grade.

Set tiered tasks for different students

Here’s the move that separates targeted assignment from a one-size task: turn the difficulty dial independently for groups within the same class, on the same sub-topic.

A worked example for a mixed class studying simultaneous equations:

• Foundation group: topic = simultaneous equations, difficulty = low–medium, command words “solve” with two-step linear systems. Six questions that build confidence and secure the method.
• Secure group: same topic, difficulty = medium, including one worded problem so they apply the method, not just execute it.
• Stretch group: same topic, difficulty = high — non-linear systems, multi-step problems, an “show that” proof-style question.

Three tasks, same sub-topic, three difficulty bands — and crucially, every student is working on the right thing for them at the same time. That’s the same effort from each student aimed where it counts, instead of the bottom third drowning and the top third bored. Running a mixed-ability room solo is genuinely hard, and where the bank’s filters help most is exactly this; there’s more on the wider problem in teaching a mixed-ability IGCSE class solo.

When you tier like this, keep the topic constant across groups even as difficulty changes. That way the whole class can be discussed together afterwards — same concept, different depth — and you’re not managing three separate topics in one room.

Scaffold across a week, not just within a task

Difficulty selection also plays out over time. A scaffolded sequence on one sub-topic might run:

1. Day 1: low-difficulty recall, whole class, to establish the floor.
2. Day 3: medium-difficulty “explain,” with the strugglers kept on a gentler set and the secure students stepped up — the marked Day-1 results tell you who goes where.
3. Day 5: high-difficulty applied questions for those ready, a consolidation repeat at medium for those who aren’t.

The difficulty ramp here lives across the week, and who gets which level is decided by data from the previous task rather than by a guess at the start. That’s the loop: assign, see the marked spread, re-tier, assign again. Selection isn’t a one-time setup — it’s a steering wheel you turn each time results come back.

A quick checklist before you assign

Before you commit any task, run these four questions:

• Right topic resolution? Sub-topic or command word, not just the broad topic — unless you genuinely want broad revision.
• Right difficulty for the moment? Does the level match where you are in teaching this — recall, consolidation, or application?
• Does it ramp? Easy-to-hard inside the task, so you get a diagnostic gradient and not a wall.
• Same for everyone, or tiered? If the class spread is wide, set two or three difficulty versions on the same topic rather than one compromise task.

If you can answer those, you’ve made the decision the tool can’t make for you — and the actual assigning takes under a minute.

The honest caveats

Selecting well is most of the battle, but a few things to stay realistic about:

• Difficulty bands are a guide, not gospel. Boards’ grade boundaries shift, and a “medium” question on an unfamiliar context can play hard. Glance at a question before you bank on its label, especially for the foundation group where a single confusing context can derail the whole task.
• Tiering has a limit. Three difficulty versions is usually the practical ceiling; beyond that you’re managing complexity that costs you more than it gains. Don’t over-engineer.
• Selection doesn’t replace teaching. A perfectly targeted set of “explain” questions tells you the gap is still there; it doesn’t reteach the concept. The assignment finds and drills — you still close the loop in the room.

How this looks in practice

If you want to select this way, Tutopiya for Teachers is built on a bank of real Cambridge and Edexcel past-paper questions filterable by topic, sub-topic, difficulty and command word — which is exactly the resolution the decisions above need. You can build a ramped task or three tiered versions on one sub-topic, assign to the whole class or to specific groups, and because every question carries its mark scheme, the results come back auto-marked with per-student and class analytics — so the data you need to re-tier the next task is already there. Question-bank access is on the free tier, enough to try this with one class. For the mechanics of pushing a task out once you’ve chosen the questions, the assign-a-quiz walkthrough covers it click by click.

FAQ

How do I choose past paper questions by topic and difficulty for a mixed class? Keep the topic constant and vary the difficulty by group: a low–medium set for the foundation students, medium for the secure middle, and high-difficulty applied questions for the stretch group — all on the same sub-topic. That way everyone works on the right level of the same concept at once, and you can discuss it as a class afterwards.

What difficulty should I assign just after teaching a topic? Low difficulty with narrow command words — “state,” “name,” “identify,” short MCQ. Right after introducing a concept you’re checking that the basics landed, not testing application. Save “analyse” and “evaluate” questions for the end of the topic when failure actually tells you something specific.

Should a single task be all one difficulty? Usually not. A ramp from easy to hard gives every student a confident start and produces a diagnostic gradient — when you see exactly where a student fell off, you learn their ceiling on that sub-topic instead of a flat pass or fail. A six-question ramp of two low, two medium, two high works well.

How do I assign past paper questions to students to target a specific weakness? Filter to the precise sub-topic and the command word that exposed the gap, not the whole topic. “They drop ‘explain’ questions on rates of reaction” becomes a six-question drill on exactly that — far more useful than reassigning the entire chapter. Use the marked results to decide whether to drill again or step the level up.

Is selecting questions by topic and difficulty free to try? Tutopiya includes question-bank access with topic, sub-topic, difficulty and command-word filters on its free teacher tier, which covers one class — enough to build ramped and tiered tasks end to end before deciding whether you want the full toolkit.

The bottom line

Assigning past paper questions well is a selection problem, not a software problem. The questions exist; the skill is turning two dials — topic and difficulty — with intent: matching the level to the moment in your teaching cycle, ramping difficulty inside a task for a clean diagnostic, filtering to the exact sub-topic that’s costing marks, and tiering the level across groups so everyone works where it counts. Get the choice right and the assigning is trivial. Get it wrong and no amount of clicking saves the task.

Build a tiered, topic-targeted task free with one class →

Ready to Excel in Your Studies?

Get personalised help from Tutopiya's expert tutors. Whether it's IGCSE, IB, A-Levels, or any other curriculum — we match you with the perfect tutor and your first session is free.

Book Your Free Trial