Cognitive scientists have spent two decades rating the study methods students actually use, and the scoreboard is lopsided. The field's benchmark review, Dunlosky et al. (2013) in Psychological Science in the Public Interest, graded 10 popular techniques and rated only 2 as high utility: practice testing and spaced practice.
The techniques most students rely on, rereading and highlighting, sit at the bottom of the same list. This guide works through 11 mainstream methods and gives each a verdict: works, wastes time, or depends on execution.
What Makes a Study Method Effective?
One criterion separates the winners from the rest: performance after a delay, not performance five minutes after studying. Most methods look fine immediately. The rankings only become visible a week later.
Roediger and Karpicke (2006) showed the crossover directly. Students who restudied a passage beat students who took a practice test when quizzed 5 minutes later, 81% to 75%. One week later the result flipped: 56% for the tested group, 42% for the restudy group.
Confidence tracked the wrong signal. The repeated-study group rated its own future recall highest, then performed worst. Researchers call this the fluency illusion.
The illusion shapes behavior at scale. When Karpicke, Butler, and Roediger (2009) surveyed 177 college students, 84% listed rereading as a study strategy and 55% ranked it first. Roughly 11% reported testing themselves. A century of learning science points one way; habits point the other.
The table below applies that delayed-retention standard to all 11 methods.
Study Methods That Actually Work
1. Active Recall (Retrieval Practice)
Close the materials and pull the information out of memory: flashcards, past-paper questions, blurting onto a blank page, answering end-of-chapter problems cold. The retrieval attempt itself is what builds the memory.
The evidence is the strongest in the field. In Roediger and Karpicke's second experiment, students who read a passage once and took three recall tests remembered 61% a week later. Students who read it four times remembered 40%, despite averaging 14.2 read-throughs against 3.4.
The Verdict: works, in any subject. AskSia's Mock Exam mode runs the same principle in real exam format, with auto-graded rationale on every miss.
2. Spaced Repetition
Same total study time, split into sessions separated by gaps. A quantitative review covering 317 experiments confirmed the effect across materials (Cepeda et al., 2006), and follow-up work found an optimal gap improved final recall by up to 150% against massed review (Cepeda et al., 2009).
Gap size should scale with the test date. Across more than 1,350 learners, the best gap ran 20–40% of a one-week test delay but only 5–10% of a one-year delay (Cepeda et al., 2008). A quiz next week rewards 1–3 day gaps; year-long retention rewards gaps of several weeks.
The Verdict: Works, and it is the exact opposite of cramming the night before.
Spacing is also the kernel inside every viral revision rule on study TikTok. The decoder below checks four of them against the data.
3. Interleaving
Mix problem types inside one session instead of finishing a block of each type. In a nine-week classroom experiment with 140 students, interleaved practice produced 72% on an unannounced test two weeks later against 38% for blocked practice, an effect of d = 1.05 (Rohrer et al., 2014).
It feels worse while working. Taylor and Rohrer (2010) found interleaving lowered practice-session accuracy yet doubled scores on a test the next day.
The Verdict: works for problem-based courses such as calculus, physics, and accounting. It has little to offer a pure reading list.
4. The Feynman Technique
Explain the concept in plain language as if teaching it, find the point where the explanation stalls, relearn that piece, and simplify again.
The branded technique has no trial base of its own, but its ingredients do. Self-explanation earned a moderate-utility rating in the 2013 review, and explaining from memory is retrieval practice by another name.
The Verdict: works, with one condition. The explanation must happen with notes closed. Reading your notes aloud is rereading in costume.
Study Methods That Does Not Work
1. Rereading
Rereading is the default. In the 2009 survey of 177 students, 84% listed it and 55% called it their number-one strategy. It earned a low-utility rating anyway: the second pass adds little the first pass did not.
The comparison below shows what the habit costs at a one-week delay.
Same passage, same students, same week (Roediger & Karpicke, 2006, Experiment 2). Four times the exposure bought 21 points less retention.
The Verdict: read once, then switch to retrieval. Rereads are for reference, not for learning.
2. Highlighting
Marking text requires no processing of it. The 2013 review found no reliable benefit, with occasional costs on questions that required connecting ideas across a passage.
The Verdict: skip as a study method. A highlight is a bookmark for later retrieval work, nothing more.
3. Summarization
Writing summaries rated low utility for untrained students in the same review; benefits appeared only after explicit instruction in how to summarize. Copying key sentences with the book open is transcription, not learning.
If condensed material is what you need, let a tool produce it. AskSia's Sia Note compresses a 200-page chapter into one worked note, so your minutes go to retrieving from the note rather than manufacturing it.
The Verdict: skip as practice. Closed-book summarizing is the exception, because that is retrieval.
Study Methods That Depend on How You Use Them
1. The Pomodoro Technique
Twenty-five minutes on, five off. No study of Pomodoro measures delayed retention, because it is not a memory technique. It manages attention and lowers the cost of starting.
The Verdict: conditional. It earns its place for staying focused during long sessions and as a procrastination fix, and it inherits the value of whatever you run inside the blocks.
2. Cornell Notes
The three-zone page format has thin evidence as a format. Its active ingredient is the cue column: questions written during class that you answer later with the notes covered.
The Verdict: conditional. Quiz from the cue column and it becomes retrieval practice. Recopy notes into the layout and nothing changes.
3. Mind Mapping and Concept Maps
Karpicke and Blunt (2011) put this to a hard test in Science: one session of retrieval practice beat elaborative concept mapping on a test a week later, even when the final test required drawing a map.
The variable is the open book. Mapping while reading is decoration; mapping from memory is retrieval with structure.
The Verdict: conditional. Draw the map closed-book first, then check it against AskSia's Concept Map of the course to see which branches you dropped.
4. Study Groups
Explaining a problem to a peer is retrieval plus elaboration, and it works. Group rereading of slides inherits rereading's record. Kornell and Bjork (2007) found only 18% of surveyed students treat testing as a way to learn, which is exactly what a group can fix.
The Verdict: conditional. Quiz each other; do not co-read.
How to Match Study Methods to Courses?
The verdicts converge on a two-part core: retrieval practice builds the memory, spacing keeps it. The remaining choice is which supporting method fits the course in front of you.
Problem-based courses reward interleaved problem sets. Memory-heavy courses reward spaced flashcards. Reading-heavy courses reward closed-book recall after every reading, with a Feynman pass for concepts that refuse to compress.
Running the pairing is the hard part, because both core methods need the same materials in one place. AskSia is built as one study workspace for that reason: lectures become transcripts, transcripts become notes, notes become flashcard decks and mock exams, and retrieval plus spacing run on a single set of sources.
Method choice also shifts with the calendar. This guide covers the semester system; once an exam sits inside two weeks, switch to the compressed sequence in our exam-window study plan.
Frequently Asked Questions
What's the best method for studying?
Practice testing, also called retrieval practice, holds the strongest record. It was one of only 2 techniques out of 10 rated high utility in Dunlosky et al.'s 2013 review, with benefits across ages, materials, and test formats. The margin is large: students who read a passage once and took three recall tests retained 61% after a week, while students who read it four times retained 40% (Roediger & Karpicke, 2006). Recognition-style review, scanning notes and nodding along, does not count; the effort of recall is the mechanism. Pair it with spacing, the other high-utility technique, and the two cover most of what the evidence supports. Start by pulling past papers or writing five closed-book questions after each lecture; for standardized exams, the drills in AskSia's test prep hub are already built in retrieval format.
What is the 1 3 5 7 rule in studying?
The 1-3-5-7 rule is a pop version of spaced repetition: after learning something, review it 1, 3, 5, and 7 days later, with each pass done as self-testing rather than rereading. A 2-3-5-7 variant adds an earlier review, and a reversed 7-3-2-1 version counts down to an exam date. The fixed intervals are a crude but reasonable fit for short horizons. Across more than 1,350 learners, the optimal gap between reviews ran 20–40% of a one-week test delay, which day-scale gaps approximate well (Cepeda et al., 2008). Past roughly two weeks, the same data say to stretch gaps toward weeks rather than days, so the rule under-spaces long-range material. Hand the bookkeeping to one of the study apps we've reviewed rather than a paper calendar; tracking dates manually is what kills the habit.
What is the 9 8 7 rule for studying?
Unlike the 1-3-5-7 rule, the 9-8-7 rule has no consistent definition. Sources promoting it contradict each other: one education publisher describes it as 9 minutes of preparation, 8 minutes of study, and 7 minutes of review on one page, and as 9 hours of study, 8 hours of sleep, and 7 hours of everything else on another. No university or peer-reviewed source defines it at all. The one number in it worth keeping is the 8 hours of sleep. In a monitored challenge across two courses (N = 34), students who averaged at least 8.0 hours of sleep during finals week scored higher on the final exam even after controlling for pre-final grades, and fewer than 10% of students normally sleep that much before exams (Scullin, 2019, Teaching of Psychology). Skip the rule; keep the sleep.
How do Harvard students study?
There is no proprietary Harvard method. The university's own guidance teaches the two high-utility techniques covered above. Harvard Summer School's study advice recommends spaced repetition and practice testing by name and cites research that cramming produces short-term performance rather than durable learning. Harvard's Academic Resource Center backs it with academic coaching, peer tutoring, and note-taking systems, plus one unusually specific tip: if you speed up recorded lectures, stay under 25% compression, because measured learning drops somewhere between 25% and 50% speed-up. The honest reading is that elite institutions brand the standard evidence, not a secret alternative to it. Copy the system rather than the name: spaced review on a calendar, self-testing after every lecture, and a study group that quizzes instead of co-reads.
How many hours a day should you study effectively?
The U.S. federal credit-hour definition assumes a minimum of 2 hours of out-of-class work per credit per week, which puts a 12–15 credit load at 24–30 weekly study hours, roughly 3.5–4.5 hours on a typical day. Treat that as a ceiling estimate rather than a target, because distribution predicts retention better than duration. The spacing literature is consistent that the same 5 hours spread across five days beat 5 hours in one sitting, and optimal gaps scale with how far away the test is (Cepeda et al., 2006, 2008). Hours also stop counting when they come out of sleep; the finals-week data above showed students averaging 8 or more hours outperforming shorter sleepers. Plan in sessions per course per week, two to four short ones each, and let the weekly total land where it lands.
One honest limit. The meta-analysis behind these rankings, 242 studies and 169,179 participants with a mean effect of d = 0.56, measured mostly factual and surface-level outcomes (Donoghue & Hattie, 2021). For deep synthesis across a discipline the evidence base is thinner, and the verdicts above are strong defaults rather than laws.
