Blocked practice feels like mastery; interleaving feels like confusion — and often predicts exam transfer better. Here’s when to use each without wrecking first-time encoding.
If you study all of Topic A on Monday and all of Topic B on Tuesday, Monday will feel great. You will know “what we’re doing today.” Exams rarely extend that courtesy. They mix. They force a hidden skill: which idea applies here?
Interleaving — mixing practice of different topics or problem types — often improves later discrimination and transfer even while it lowers in-session fluency. Blocked practice keeps one type in a row and boosts short-term performance. Robert Bjork’s desirable difficulties framing predicts the emotional trap: students prefer the method that feels better now.
Rohrer and colleagues’ work on interleaved mathematics practice is a useful flagship finding for students: mixing problem types can beat blocking for later tests that require choosing a strategy.
What each method trains
| Method | During practice | Later payoff |
|---|---|---|
| Blocked | High fluency, low choice demand | Good for early encoding drills |
| Interleaved | Lower fluency, high choice demand | Better discrimination / transfer |
| Full papers | Natural interleave + time | Exam-like transfer + stamina |
| Blocked re-read of markschemes | Recognition comfort | Weak transfer |
Interleaving ≠ chaos
Mix related topics that are easy to confuse — not twenty unrelated subjects in a blender. The goal is discrimination, not randomness cosplay.
The discrimination skill exams hide
Most papers do not announce “this is a Topic 4 question.” They present a surface that could belong to two near-twins. Blocked practice trains you to execute a method once the method is given. Interleaving trains you to select the method when the surface is ambiguous.
That selection step is why green chapter decks can still lose to mixed sets. You practiced the dance with the partner labeled. The exam removes the label.
When blocking wins
- First encounter with a schema (reduce load; Sweller-aware)
- Drilling a single fragile procedure after a worked example
- Short repair on one misconception before you re-enter the mix
- Early vocabulary / part-naming before mechanism practice
Blocking is a scaffold. Fade it. If blocking is still your default three weeks into a topic, you are protecting fluency at the expense of transfer.
When interleaving wins
- After teach-back is possible on each component
- When exams mix topics (almost always)
- When you confuse near-twins
- When cards are green but papers punish cue changes
- When error logs show “used wrong method / confused X with Y”
Blocked comfort addiction
If every session needs a single chapter label to feel okay, you are training a classroom condition the exam will not provide. Add mixed sets on purpose.
A hybrid week for exam courses
- 2 blocked encode/repair sessions on focus items
- Daily retrieval from a mixed deck
- 1 deliberate interleaved question block
- Full paper or timed section as phase allows (free interleaving)
After papers, use the error log: discrimination misses become comparison notes + interleaved re-tests — not another pure block of the same chapter only.
| Week signal | Lean toward | Example move |
|---|---|---|
| New schema this week | More blocking early | Worked example → 5 similar → then mix |
| Near-twin confusions | More interleaving | Shuffle A/B items; force “which?” pause |
| Paper discrimination errors | Interleave + compare notes | Two-column differences + mixed set |
| Overwhelmed / load crash | Temporary block + clarify | Encode, then re-enter mix |
Designing interleaved sets
- Pick 2–4 topics that share surface features
- Shuffle question types
- Force a “which method?” pause before solving
- Review by concept clusters after, not only by score
- Re-interleave a week later
Optional upgrade: write the discrimination rule in one sentence after the set (“If you see X cue, prefer method A because…”). That sentence becomes a card worth keeping.
Mocks already interleave — use them
Sitting papers without studying the discrimination errors wastes the mix. Log “confused X with Y” explicitly. That line is interleaving’s syllabus.
Spacing still matters
Interleaving is not a substitute for spaced returns. Mix within sessions and still revisit across days. Discrimination without spacing fades like any other skill.
Practice the mix exams use
Sukrat Mock Exams provide authentic interleaving under time — pair with Planner mixed blocks and Flashcards that don’t live in permanent chapter silos. Product page: /products/features/mock-exams
A 40-minute interleaved block you can reuse
- Warm retrieve two near-twins (5m)
- Shuffle six mixed items (20m) � pause to name the method before solving
- Score and tag discrimination misses (5m)
- Write one comparison rule + schedule a return mix (10m)
Do this once mid-week even when content pressure is high. Recognition work will try to steal the slot because it feels calmer. Protect the mix anyway � that is the skill papers bill you for.
If accuracy collapses, shrink to two topics, repair models, then remix. Interleaving without a schema is not desirable difficulty; it is noise.
FAQ
Won’t interleaving slow my content coverage?
It can slow short-term coverage and speed long-term usability. Coverage without discrimination is fragile. Better: cover enough to encode, then mix practice.
Is shuffling flashcards enough?
Necessary baseline; add mixed exam-style items for fuller transfer. Cards that never leave chapter tags still train soft blocking.
What about subjects with strict sequences?
Keep prerequisites ordered for encoding; interleave practice after the chain exists. Sequence for schema; mix for selection.
How mixed is too mixed?
If accuracy collapses to guessing with no feedback learning, reduce to 2 near-twins, repair models, then remix. Confusion without correction is just noise.
Should every homework set be interleaved?
No. Early homework after a worked example can stay blocked briefly. Fade toward mix as fluency rises.
Related reading
- Interleaving on Your Calendar
- Full Papers Train Transfer
- Exam Transfer vs Recognition
- When Flashcards Aren’t Enough
- Error Logs After Mock Exams
- Schedule the Struggle
Keep going across Sukrat
Mix what the exam will mix
Blocked practice builds early fluency. Interleaving builds the choice skill hiding inside most papers: which idea is this?
Encode enough to have models. Then mix retrieval and questions. Let mocks surface discrimination errors — and repair those on purpose.
If practice never feels jumpy, transfer may feel unfair later — even when the board was being predictable.