The Science of How We Learn: What Research Says About Studying

The Most Popular Study Methods Are Among the Least Effective

In 2013, psychologists John Dunlosky and colleagues at Kent State University published a landmark review in Psychological Science in the Public Interest that evaluated ten commonly used study techniques against the accumulated evidence from cognitive research. The results were striking. Highlighting and underlining — the most widely used study methods — were rated "low utility." Rereading and summarization received the same low rating. The highest-rated techniques — spaced practice and retrieval practice — were the methods students used least often. The gap between what students believe works and what research demonstrates works is large, persistent, and consequential.

Why do students cling to ineffective methods? Partly because effective techniques feel harder. When you reread a chapter you've already read, the words feel familiar — fluency creates a sense of mastery. Cognitive scientists call this fluency illusion: the ease of processing a familiar text is mistaken for understanding. In contrast, when you close the book and try to recall what you read — which is harder and produces more errors — it feels like failure. But those errors, and the effort of correcting them, are where durable learning actually happens.

The Evidence Hierarchy for Study Techniques

Spaced Repetition: Fighting the Forgetting Curve

Hermann Ebbinghaus, a German psychologist working in the 1880s, discovered the forgetting curve by systematically memorizing and testing nonsense syllables. His data showed that without review, memories decay in a predictable pattern — roughly 50% of new information is lost within an hour, and most of the remainder within a day. But he also discovered the spacing effect: the same number of repetitions spread over time produced far better long-term retention than the same repetitions massed together.

Ebbinghaus's core insight — that memories need to be revisited just before they are forgotten to be consolidated most efficiently — forms the basis of modern spaced repetition systems. Software like Anki uses algorithms derived from SuperMemo, developed by Polish researcher Piotr Wozniak in the late 1980s, to schedule reviews at optimally widening intervals. A card first reviewed after one day might next appear after three days, then ten days, then thirty, as the memory strengthens. Studies show spaced repetition produces retention rates that massively outperform cramming over long time periods.

Retrieval Practice: The Testing Effect

Robert Bjork at UCLA introduced the concept of desirable difficulties — the counterintuitive idea that introducing obstacles to learning can improve long-term retention. Testing is the most powerful example. A 2006 study by Henry Roediger and Jeffrey Karpicke, published in Science, compared students who studied material four times against students who studied once and then took three retrieval tests. One week later, the test group recalled 61% of the material; the study-only group recalled 40%.

• The benefit of retrieval practice holds even when the tests are graded — the act of retrieval, regardless of success, strengthens the memory trace
• Retrieval practice works across subjects — foreign language vocabulary, scientific concepts, historical dates, and mathematical procedures all show the testing effect
• Feedback after retrieval matters: correcting errors immediately after a test prevents the wrong answer from consolidating

Interleaving: The Confusion That Helps

Blocked practice — doing twenty problems of type A, then twenty of type B, then twenty of type C — feels efficient and produces high performance within the practice session. Interleaved practice — randomly mixing problem types — produces more errors during practice but dramatically better performance on subsequent tests. A 2010 study by Rohrer and Taylor showed that students who used interleaved math practice scored 43% higher on a delayed test than students who used blocked practice, despite scoring lower during the practice itself.

The mechanism appears to be discrimination learning: when problem types are mixed, the solver cannot rely on the current context to identify which procedure applies. They must identify the type of problem — an additional cognitive step that is effortful during practice but produces more flexible, transferable knowledge. Many teachers and students avoid interleaving precisely because the short-term performance dip feels like failure. Recognizing the long-term benefit requires trusting research over subjective experience.

Sleep, Consolidation, and Cognitive Load

Memory consolidation happens largely during sleep. During slow-wave sleep, the hippocampus replays newly acquired memories, strengthening their transfer to cortical long-term storage. During REM sleep, procedural and emotional memories undergo additional processing. A 2003 study at Harvard Medical School found that participants who slept after learning a motor sequence showed 20–30% improvement in performance the next morning, while those who stayed awake showed none. The implication for students who sacrifice sleep to cram is direct: they are eliminating the biological process that converts study into lasting memory.

What This Means for Teaching

The mismatch between evidence-based pedagogy and classroom practice is not limited to students. Many teaching methods — lecturing, assigning readings to be reviewed before class, giving infrequent high-stakes exams — are structured in ways that minimize retrieval practice and spaced repetition, precisely the techniques that research most strongly supports. The flipped classroom model, low-stakes frequent quizzing, mixed problem sets, and explicit sleep hygiene guidance all have empirical support. The barrier to implementation is rarely information — most educators have heard of spaced repetition and retrieval practice. The barrier is institutional inertia, time constraints, and the difficulty of redesigning curricula around principles that feel counterintuitive until the evidence is internalized.