In a 2006 Washington University experiment, students who read a science passage four times recalled 83% of it five minutes later. A week on, that same group could produce only 40%. A second group spent most of their sessions testing themselves instead of rereading. After a week, they held 61%.
The students who studied hardest remembered the least. This is the central problem with how most people prepare for exams. The techniques that feel productive and the techniques that work are not the same list.
Cognitive scientists have ranked the common ones. Two come out on top. Most students use neither.
Why Does Rereading Fail on Exam Day?
Rereading and highlighting build recognition, not recall. Reread a sentence and your brain registers familiarity, then signals that you know it. The exam hands you a blank page, where familiarity is worthless. Recall is the only thing scored.
The 2006 study caught this gap in numbers.
Five minutes after studying, rereaders led self-testers 83% to 71% and felt sure they would remember. That confidence was the trap. By one week, rereaders had dropped to 40% while testers held 61%, a reversal driven entirely by how each group spent its time.
Reviews since then put the testing advantage in the moderate-to-large range, a Hedges' g near 0.50–0.61, and it widens as the gap to exam day grows.
10 Study Methods Actually Work
In 2013, John Dunlosky and four co-authors reviewed ten common learning techniques for Psychological Science in the Public Interest. They scored each on how well it holds up across ages, subjects, and test formats. Only two earned a high-utility rating.
The ranking embarrasses common habits. Summarizing, highlighting, the keyword mnemonic, and rereading all landed in the low tier. Those are the techniques students report using most. Practice testing and spaced study, the two at the top, are the two most people skip.
Interleaving earned a moderate rating. It means mixing problem types instead of drilling one kind in a block.
That has a practical angle. Seeing how a course's topics connect turns interleaving from random to deliberate.
AskSia's Concept Map renders a whole course as a navigable tree from foundations to synthesis, so you can pull problems from related branches in one sitting. The same cycle works on any subject, and our chemistry study guide runs it on one course end to end.
How Do You Build a Study Cycle?
A working study cycle has two jobs: force recall, and space it out. Everything else is packaging. The build below assumes a college course with weekly material and an exam four to eight weeks out.
The cycle spans tools students usually keep in separate apps: notes, flashcards, practice tests, a scheduler. AskSia folds these into one workspace, which matters here because the system fails the moment a piece lives somewhere you forget to open. Lecture capture, source Q&A, and practice generation sit together.
The procedure is five steps.
- Write notes you can be tested against. On the day you learn material, phrase the key points as questions, not paragraphs. Prose invites rereading. Questions force recall later. Condensing a dense chapter into a tight, testable one-page reference sheet does the same job.
- Build your test items on Day 0. Pull the five to ten concepts that matter most and turn each into a prompt. This is the line between a deck you can drill and a wall of text you will only skim.
- Self-test closed-book first. Every review starts with the notes shut. Recall, then check. The effort of retrieval is the part that builds memory, even when you miss.
- Space the retests. Return to each topic after a couple of days, then a few days more, then about a week. Catch the memory as it fades, not after it is gone.
- Finish under exam conditions. In the final week, sit a full practice test, timed and uninterrupted. The format is its own skill. For standardized tests the logic is identical, as our SAT study approach lays out.
One cadence for a single topic looks like the schedule below. Treat the gaps as flexible.
Two of these steps are exactly what AskSia automates. Flashcards built on FSRS scheduling handle the spacing, resurfacing each card as your recall weakens rather than on a fixed date. Mock Exam mode covers the last step, generating adaptive practice in real exam format and grading it with rationale, so a wrong answer becomes a correction.
Does the 2-3-5-7 Method Work?
The 2-3-5-7 method is the most-searched name for a spaced-review schedule, and it has a definition problem. Across published study guides, the same four digits describe at least four different timetables.
Some sites count forward from the day you learn a topic. Others count backward from the exam, a different tool for a different moment. A few relabel it 1-3-5-7 or stretch the gaps to 2, 5, 10, and 17 days.
The numbers are not the method.
What sits underneath all of them is the spacing effect, documented in a 2006 meta-analysis by Cepeda and colleagues spanning hundreds of experiments.
Fixed schedules approximate something an algorithm can do exactly. FSRS, the scheduler behind AskSia's Flashcards, sets each card's next review from your own recall history rather than a shared rule. Miss a card and it returns sooner. Know it cold and it drifts further out. The 2-3-5-7 method is a hand-tuned guess at that same curve.
For one subject over a week, the manual version works well enough. It breaks across five courses at once, when the review sessions stack until tracking them becomes its own job.
What can disrupt a study plan?
Three failure modes account for most collapsed study plans. None of them is laziness.
The first is the rereading illusion: hours that feel like work but build only familiarity. The fix is mechanical. Notes closed and recalling, it counts. Notes open and reading, it does not.
The second is cramming. Massed study the night before produces the same short spike the 2006 rereaders saw: strong at five minutes, gone in a week. It can pass an exam you forget by the next unit, a poor trade when later material builds on earlier. For genuine time crunches, our day-before memorization tactics cover what to prioritize, but treat them as triage.
The third is tracking. Spaced review creates a scheduling load that grows every week, and most students quit the system not from weak effort but because hand-tracking what to review, on which day, across several courses is genuinely hard.
Focus failures sit under all three. The honest fix is environmental, not motivational: put the phone in another room, work in fixed blocks, and start before you feel ready.
Pomodoro-style sessions often run 25 minutes of work to a 5-minute break.
When one concept is the thing stalling you, hearing it explained another way clears the block faster than willpower. That is what AskSia's AI Tutor does, working a problem several ways until one lands. Our guide to breaking the procrastination loop goes deeper on the triggers.
Frequently Asked Questions
What is the best study method for exams?
Two methods, per the 2013 Dunlosky review: practice testing and distributed practice, the only two of ten common techniques rated high-utility. Practice testing means self-quizzing with notes closed. Distributed practice means spacing sessions over days instead of massing them. In the 2006 Roediger-Karpicke study, self-testers held 61% of a passage after a week against 40% for rereaders. Both beat popular habits like highlighting and summarizing, which the same review rated low-utility. Combine them: turn each topic into questions, test yourself closed-book, then return on a spaced schedule. Build the question bank as Flashcards in AskSia so the spacing runs on the FSRS algorithm instead of a calendar you maintain by hand.
What is the 2-3-5-7 study method?
It is a spaced-repetition schedule, and its definition shifts by source. The common forward version reviews a topic 2, 3, 5, and 7 days after you first learn it. An exam-countdown version flips the order, placing sessions 7, 5, 3, and 2 days before the test. Some guides relabel it 1-3-5-7 or widen the gaps to 2, 5, 10, and 17 days. All share one principle: review at expanding intervals using recall, not rereading. The specific numbers are a heuristic, not a research finding. The underlying spacing effect comes from work like the 2006 Cepeda meta-analysis of hundreds of experiments. Use the manual version for one subject over a week or two. For several courses at once, let an app set the intervals, since hand-tracking them is where the method usually fails.
What's the best way to study for an exam?
Work backward from the exam date and build recall into every session. Start the week you learn material, not the week before the test. Convert notes into questions on day one, self-test closed-book, space the retests across the following days, and sit at least one full timed practice exam before the real one. This mirrors what topped the 2013 Dunlosky ranking: testing plus spacing. It also avoids the trap the 2006 data exposed, where rereaders felt confident at 83% recall after five minutes, then fell to 40% after a week. Confidence during study is not evidence of retention. AskSia's exam-format guides and a timed Mock Exam in the final week let you practice the real format before it counts.
How can I focus 100% on my study?
Full focus is mostly an environment problem, not a willpower one. Remove the phone from the room, since its presence alone measurably cuts available attention. Work in fixed blocks with defined start and stop times rather than open-ended sessions. A common structure is 25 minutes of work to a 5-minute break, repeated four times before a longer rest. Start before you feel ready, because motivation tends to follow action rather than precede it. When attention keeps breaking on one concept, the cause is often confusion rather than distraction, and getting that idea explained a different way clears it faster than forcing concentration. Schedule breaks deliberately so rest is planned, not a crash. Treat focus as a setup you build, then protect.
Is cramming ever effective?
Cramming works for a narrow case and fails the rest. Massing all study into one session before an exam produces a real but short-lived spike. The 2006 Roediger-Karpicke data is the clearest illustration: heavy recent exposure looked strong at 83% recall after five minutes, then collapsed to 40% a week later. If an exam is tomorrow and the material is self-contained, a focused cram can carry you through it. The cost arrives later, in any course where the next unit assumes you kept the last. For cumulative subjects, the spaced cycle is not optional polish. It is the only version that survives to the final exam. If you are truly out of time, prioritize the highest-weight topics and the concepts other topics depend on, and accept that this is triage rather than a study plan.
How many hours should you study for an exam?
There is no fixed number, because hours measure time, not retention, and the two diverge sharply. A common planning baseline is two to three hours of study per credit hour each week, so a three-credit course implies roughly six to nine hours weekly across the term. The 2006 and 2013 research both show that how you spend the hours matters more than how many: a tested, spaced hour outperforms several reread ones. Front-load steady weekly review instead of banking one long session near the exam. Track effective study, meaning time spent in active recall, rather than total time with the book open. Map the week's topics in AskSia's Concept Map to see which need more passes, then weight your hours toward the weak branches.
