Interleaved vs. Blocked Practice

Practice Schedule Design

Interleaving 50%

Trials per problem type 15

Retention delay 7 days

Problem Types

Live Readouts

Practice accuracy

–

Avg discrimination D

0.00

Avg retrieval S

0.00

Predicted test acc.

–

Run a practice session, then press "Test After Delay" to compare blocked vs. interleaved retention.

How it works

This models the contextual-interference / interleaving effect from instructional-design research (Rohrer & Taylor 2007; Bjork's "desirable difficulties"), using the classic example: four types of geometry volume problems (prism, cone, wedge, spheroid) practiced either blocked (AAAA BBBB CCCC DDDD) or interleaved (ABCD ABCD…).

Each trial updates two per-category quantities toward a ceiling of 5, with diminishing returns:

repeat trial (same type as last):
  S += 0.55·(1 − S/5)   D += 0.12·(1 − D/5)
switch trial (different type than last):
  S += 0.35·(1 − S/5)   D += 0.55·(1 − D/5)

S is surface retrieval strength — it grows fastest under blocked, low-effort repetition. D is discrimination strength — how well the learner can tell problem types apart and pick the right method — and it only grows when the brain must switch strategies mid-session, which is exactly what interleaving forces.

Immediate practice accuracy is p = σ((S − 2.5)/1), so blocked practice looks better while it's happening. But retention decays exponentially like a classic Ebbinghaus forgetting curve, and the decay time-constant itself depends on D:

τ = 5 · (1 + 1.4·D)
S_after_delay = S · e^(−days / τ)
test accuracy = σ((S_after_delay − 2.5)/1)

Higher D means slower decay — durable, transferable learning — which is why interleaved practice, despite lower accuracy during the session itself, typically wins on a delayed test. Slide "Interleaving" to 0% and 100% and compare the predicted test accuracy to see the effect directly.

Retrieval strength (S) — solid bar Discrimination (D) — thin sleeve

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.

Interleaved vs. Blocked Practice: An Instructional Design Simulator