Psychology › Thinking, memory & biases › Memory and how it works
Spacing and retrieval practice
1885; reviewed 2013· Finding or experiment
Well replicatedStrong, repeated evidence.
Spreading study over time and testing yourself, rather than rereading, are the two best-supported ways to make learning stick.
See it on the Psychology map →What it says
Ebbinghaus noticed in 1885 that repetitions spread over days beat cramming; hundreds of later studies confirm this 'spacing effect'. Retrieval practice, pulling answers from memory with flashcards or quizzes, beats rereading even though it feels harder (Roediger and Karpicke, 2006). A 2013 review led by John Dunlosky rated ten common study techniques: only these two earned 'high utility'. Highlighting, rereading and summarizing rated low.
Why it matters
They help across ages and subjects, from vocabulary to professional licensing exams, cost nothing, and remain underused because rereading feels more productive.
Think of it like
Like lifting weights: the effort of hauling an answer back out of memory is what builds the strength.
The evidence
Well replicated. Hundreds of lab and classroom studies, meta-analyses by Nicholas Cepeda and colleagues (2006) on spacing and Christopher Rowland (2014) on testing, and the 2013 Dunlosky review.
Open questions
The best gap between sessions for a given goal, and how well lab gains carry over to complex real-world skills, are still being worked out.
Inside this (2)
Connections (5)
▸Sleep and dreams boost the payoff of Spacing and retrieval practiceWell replicated
The connection
Spacing study across days may work partly because sleep falls between sessions. During sleep, especially deep slow-wave sleep, the brain replays and stabilizes newly learned material. Studying, sleeping, then reviewing beats the same study crammed into one day. The role of dreaming itself is far less clear.
Who showed it and when
Lab studies from John Jenkins and Karl Dallenbach at Cornell (1924) to Stéphanie Mazza's 2016 study in Lyon of sleep between learning sessions.
The evidence
Controlled comparisons of sleep versus wake intervals, recordings of memory replay in rodents and humans, and sleep-deprivation studies showing weaker retention.
Open questions
Whether sleep actively strengthens memories or mainly shields them from interference is debated, and the size of the benefit varies across studies.
▸Neuroplasticity: the brain rewires with use underlies Spacing and retrieval practiceWell replicated
The connection
Learning physically changes the brain: connections between neurons that fire together strengthen, as Donald Hebb proposed in 1949. Spacing and retrieval practice appear to work by repeatedly reactivating these connections, which builds more durable changes than one long session. The brain-level explanation is partly inferred from animal studies.
Who showed it and when
Hebb at McGill, 1949; long-term strengthening of synapses found by Timothy Bliss and Terje Lømo in Oslo, 1973; Eric Kandel's sea-slug studies, Nobel 2000.
The evidence
Animal studies show spaced training produces longer-lasting synaptic changes than massed training; human evidence is mostly indirect, from brain imaging.
Open questions
Exactly which cellular changes make spaced learning last longer in humans, and how synapse-level findings scale up to real study habits.
▸Spacing and retrieval practice exploit the mechanics of Memory and how it worksWell replicated
The connection
Spacing and testing work with memory's known quirks. Some forgetting between sessions makes the next retrieval harder, and effortful retrieval strengthens a memory more, what Robert Bjork called 'desirable difficulties'. Each successful recall also makes the memory easier to find next time.
Who showed it and when
Ebbinghaus in Berlin, 1885; Robert and Elizabeth Bjork at UCLA from the 1990s; Henry Roediger at Washington University in St. Louis.
The evidence
Controlled experiments show testing beats restudying even without feedback, with benefits that grow as the delay before the final test lengthens.
Open questions
The exact mechanism is debated: retrieval may add new routes to a memory, strengthen existing ones, or update the context it is linked to.
▸Learning styles (visual, auditory, kinesthetic) lack the evidence behind Spacing and retrieval practiceWell replicated
The connection
Both promise better learning, but only one delivers. Matching lessons to a 'visual' or 'auditory' style shows no benefit in controlled tests. Spacing study sessions out and quizzing yourself instead of rereading improve retention for nearly everyone, often substantially. Time spent diagnosing styles could go into techniques that work.
Who showed it and when
Harold Pashler and colleagues' 2008 review debunked style matching; John Dunlosky's 2013 review of ten study techniques rated practice testing and spaced practice highest.
The evidence
Hundreds of controlled experiments support spacing and retrieval practice; properly designed style-matching studies, such as Rogowsky and colleagues' in 2015, find no benefit.
Open questions
Why schools still teach learning styles while underusing spacing and self-testing; surveys find most teachers believe in styles.
▸Deliberate practice and the '10,000-hour rule' shares principles with Spacing and retrieval practiceMixed evidence
The connection
Deliberate practice and good study techniques share a core idea: effortful practice beats easy repetition. Ericsson's recipe of working at the edge of ability, getting quick feedback and fixing specific weaknesses overlaps with memory research showing that self-testing and spaced practice make learning stick, even though they feel harder. Both reject passive rereading and mindless hours.
Who showed it and when
Ericsson's 1993 violin study in Berlin; Robert Bjork's 'desirable difficulties' work at UCLA from 1994; John Dunlosky's 2013 review of study techniques.
How solid is it?
The overlap is conceptual: the two research lines developed separately, but each finds effortful, feedback-rich practice works better than comfortable repetition.
Open questions
Whether lab memory findings explain expert skill, which also involves perception, motor control and talent; the two fields have rarely been tested together.