The Spacing Effect: Why Spreading Study Out Beats Cramming

If you take one idea from a century of memory research into your exam prep, make it this: the same amount of study time produces far more durable memory when it's spread out than when it's crammed into one block. That's the spacing effect, and it's about as close to a law as learning science gets. This post is the deep dive into the actual evidence — what was measured, how big the effect is, and how far apart your reviews should be.

A 140-year-old discovery — replicated in 2015

In the 1880s, Hermann Ebbinghaus tested his own memory on lists of nonsense syllables and plotted how fast he forgot them — the famous forgetting curve. His savings dropped to roughly a third within a day and about a fifth within a month. He also noticed something more useful: when he spread repetitions across days instead of massing them back-to-back, he needed fewer total repetitions to reach the same recall. Spacing wasn't just better; it was more efficient.

A single self-experiment from 1885 would be a shaky foundation on its own — which is why it matters that Murre and Dros (2015) replicated it directly: 70 hours of learning over 75 days, modern controls, and a forgetting curve that matched Ebbinghaus's original point for point. One intriguing detail from the replication: recall showed a small upward bump around the 24-hour mark, consistent with sleep actively consolidating the memory overnight.

How big is the effect? The meta-analysis

The definitive quantitative answer comes from Cepeda et al. (2006), who synthesised 317 experiments across 184 articles on distributed practice. Across 271 direct comparisons of spaced versus massed practice, final recall averaged 47% for spaced practice versus 37% for massed practice — and only 12 of the 271 comparisons failed to favour spacing. In a century of experiments across ages, materials and delays, spacing wins almost every time. Effects of that consistency are vanishingly rare in psychology.

Single studies show the same thing more vividly. In Rohrer and Taylor (2007), students who spread their maths practice across two sessions scored 74% on a later test; students who massed the identical problems into one session scored 49%. Same problems, same total work — the only difference was the calendar.

It's also why Dunlosky et al.'s (2013) landmark review of ten study techniques rated distributed practice one of only two with high utility, alongside practice testing. (Full rankings: which study techniques actually work.)

How far apart should reviews be?

The most practical study here is Cepeda et al. (2008), which tested 1,354 people across gaps of up to 3.5 months and final tests up to a year away. The result — the "temporal ridgeline" — is that the best gap depends on how far away your test is, and it's a sliding ratio, not a fixed number:

Two takeaways. First, for exams months away — NEET, JEE, UPSC — your reviews should be much further apart than feels comfortable. Reviewing tomorrow feels safer; reviewing next week works better. Second, a slightly-too-long gap costs you far less than a too-short one, so when in doubt, space wider.

In practice you shouldn't schedule any of this by hand: this ratio logic (and more) is what spaced-repetition algorithms compute per fact, automatically.

Why spacing works

The leading explanation is beautifully simple: retrieval is hardest, and therefore most useful, when you've started to forget.

• Cram three reviews into ten minutes and the second and third are trivial — the answer is still echoing in your head, so your brain does almost no work.
• Space those three reviews across a week and each one catches you just as the memory is fading. Pulling it back from the edge of forgetting is effortful, and that effort is exactly what strengthens the memory.

This is why spacing pairs so naturally with active recall: the spacing creates the difficulty, and the act of retrieving is what does the work. Bjork and Bjork (2011) put both under one name — desirable difficulties: conditions that make practice feel worse while making learning demonstrably better. Sleep adds a second mechanism: memories consolidate between sessions, so multiple sessions give consolidation multiple passes at the material.

How much practice is enough?

Spacing tells you when to review; Rawson and Dunlosky (2011) asked how much. Across three experiments with 533 students, the most efficient durable schedule was: practise until you've recalled each item correctly three times, then relearn it in about three widely-spaced later sessions. More initial drilling than that bought almost nothing. The prescription — a few honest recalls now, a few spaced returns later — is essentially what a good flashcard algorithm implements for you.

Why cramming feels like it works (but doesn't)

Cramming the night before genuinely boosts your recall tomorrow — massed practice wins on very short delays. The trap is that this memory is built on sand and collapses within days (the full anatomy of why cramming fails). For NEET, JEE or UPSC, where you must hold a year of material simultaneously, the maths is brutal: what you cram in October is gone by January. The spacing effect is the only realistic way to keep a vast syllabus alive.

Do this in StudyTab

The spacing effect is the why; here's the working system:

• Let FSRS pick every gap. StudyTab schedules each card with FSRS or SM-2, configurable per deck — each fact returns on the day you're about to forget that specific fact, which is the ridgeline logic applied per card instead of per subject.
• Choose your retention, see the cost. Retention presets (e.g. 85% vs 95%) let you trade daily workload against safety margin, and the workload forecast shows your review load through exam season before you commit.
• Make the cards cheap to create. Spacing only helps material that's in the system — the AI flashcard generator converts PDFs, lectures and notes into cards in minutes, so the spaced schedule covers your whole syllabus rather than the one chapter you had time to card by hand.
• Study a topic, then let it go. Trust the schedule to resurface it. The urge to re-review tomorrow is the fluency illusion talking; the algorithm is applying Cepeda's ratios for you.

Bottom line

Cramming buys you tomorrow and bankrupts you by next month. Spacing the same hours across days costs nothing extra, roughly moves you from 37% to 47% recall on the century of evidence, and compounds further when the reviews are retrievals. Study it, sleep on it, and review it just as it's slipping — that rhythm is the spacing effect, and it's the closest thing to a cheat code that studying has.