You’ve probably had this experience. You spend an evening with a textbook, highlighter in hand, and by the end the pages glow yellow and the material feels familiar. Then the test arrives and half of it is gone.
That isn’t a character flaw. It’s a method problem. Most of us were never taught how learning works, so we fall back on whatever feels productive, and the stuff that feels productive is often the stuff that works least.
Learning researchers have spent decades testing which habits actually move results. A lot of the answers turn out to be cheap and a bit counterintuitive. Below are 18 of them. Some are about how you study, some about when, and a few about the body that does the studying. You don’t need all eighteen at once. Pick two or three, run them for a few weeks, and see what changes.
1. Close the book and pull it out of memory
Rereading is comfortable. It’s also the default for a lot of people. In a November 2024 survey published in Frontiers in Education, Sameer Aljabri asked 202 English majors at Umm al-Qura University in Saudi Arabia how they study. Highlighting or underlining came out on top at 38.3% of open-ended answers. Self-testing on its own? Just 8.6%.
The problem is that recognising something on a page and being able to produce it are two different skills, and exams want the second one. Retrieval practice trains it directly. Jeffrey Karpicke of Purdue University, in his 2025 review chapter for the reference work Learning and Memory, summarises a meta-analysis by Rowland (2014) covering 159 studies: recall practice beat restudying with an average effect of g = 0.50, and 81% of comparisons favoured it. One honest caveat. A small January 2026 report by Kevin Sigayret and colleagues in Frontiers in Psychology failed to find the effect in online crowdsourced samples, which the authors put down to disengaged participants rather than a flaw in the method itself.
Your move: after each section you read, shut the book and write down everything you remember on a blank page. Then check what you missed.
2. Spread your sessions out instead of cramming
Cramming works for tomorrow. It’s terrible for next month.
A meta-analysis by Ewan Murray, Aidan Horner and Silke Göbel, published in July 2025 in Educational Psychology Review, pulled together 27 studies on mathematics learning. Spacing practice out beat massing it together with an overall effect of g = 0.28. When the material was taught in isolation, the effect rose to 0.43; inside a real course it was 0.24, smaller but still there. You can read the summary in the 2025 meta-analysis of spacing and retrieval practice for maths on the White Rose research repository.
Think of it like watering a plant. A bucket once a week mostly runs off; a glass every couple of days soaks in. The comparison has limits (memory isn’t soil), but it captures why the gap between sessions matters: a little forgetting forces your brain to rebuild the memory, and rebuilding is what makes it last. If you want to fit this into a busy week, a few organizing habits that free up your calendar make it far easier to protect short, repeated slots. Book three 20-minute reviews across the next week instead of one hour on Sunday night.
3. Take a test before you’ve learned anything
This one sounds backwards. Answer questions about a topic before you’ve studied it, get most of them wrong, and you’ll remember the material better later.
Steven Pan and colleagues at the National University of Singapore tested exactly that in two studies published in the Journal of Memory and Language in January 2025. With 135 undergraduates in one study and 122 online adults in the other, guessing first improved final scores by an average of 21 percentage points compared with simply reading the material, with large effects (d = 0.93 and 1.13). And it wasn’t a lucky few: 80% of the student group and 85% of the online group benefited. Intriguingly, learners with lower working memory or reasoning scores tended to gain more, though the pattern wasn’t consistent across both studies.
Why would a wrong guess help? It primes you to notice the right answer when it shows up, a bit like knowing the question before the lecture starts. If you tend to stall before starting anything new, this also doubles as a low-stakes way to stop hesitating and just begin.
A quick experiment: before your next chapter, flip to the end-of-chapter questions and answer five of them cold. Don’t look anything up.
4. Decide exactly when and where you’ll study
“I’ll study more this week” is a wish. “If it’s 7 p.m. on Tuesday, then I sit at the kitchen table with my flashcards” is a plan.
That if-then format has a name, implementation intentions, and it’s one of the best-tested ideas in motivation research. In a January 2025 meta-analysis in the European Review of Social Psychology, Paschal Sheeran, Olivia Listrom and Peter Gollwitzer looked at 642 tests of this kind of planning. Effects ranged from d = 0.27 to 0.66 across thinking, feeling and behaviour outcomes, and were bigger when plans were written as if-then statements, when people were genuinely motivated, and when they rehearsed the plan.
The plan does the remembering so you don’t have to. You’ve already decided, which takes the daily negotiation out of it. That’s also why it pairs so well with other strategies to beat procrastination.
Start small: write one if-then sentence for each study session this week, and say it out loud once.
5. Mix different problem types in one session
Textbooks usually group practice by topic. Ten fraction problems, then ten decimal problems. It feels smooth. It’s also the easier, less useful way.
Interleaving means shuffling related problem types together so you have to work out which approach each one needs. Brunmair and Richter’s 2019 meta-analysis in Psychological Bulletin is still the most complete look at this: 59 papers, 238 effect sizes and 8,466 participants. Interleaving beat blocking with a moderate effect (g = 0.42). By my rough arithmetic, that puts the average interleaver ahead of about two-thirds of blockers. For maths the effect was smaller (0.34), for telling apart painting styles larger (0.67), and for learning word lists it actually reversed. It worked best when the categories were easy to confuse.
So it’s a tool for material where the hard part is choosing the right method, which describes most maths, physics and grammar. For your next practice set, pull problems from three different chapters and shuffle the order.
6. Plan, check and review your own learning
Strong learners don’t just study. They watch themselves study.
The Education Endowment Foundation’s Teaching and Learning Toolkit, a widely used UK evidence summary for schools, rates “metacognition and self-regulation” as one of its highest-impact approaches: an average of about seven extra months of progress, at very low cost, based on 246 studies. The Toolkit breaks it into three moves: plan how you’ll tackle a task, monitor how it’s going while you work, and evaluate what worked afterward. It notes one caveat worth knowing. Many of the studies weren’t independently evaluated, and evaluations run by people linked to a programme tend to report bigger gains.
In practice this is a small habit of noticing. Am I actually understanding this paragraph, or just moving my eyes across it? It’s the study version of learning to stop living on autopilot.
Next time, do this: before you start, write one line on what you’re trying to learn. Halfway through, rate your understanding from 1 to 5. At the end, write one sentence on what you’d change next time.
7. Explain the “why” to yourself as you go
Reading that the heart has four chambers is one thing. Asking yourself why it needs four, and answering, is another.
That process is called self-explanation, and it has solid backing. A 2018 meta-analysis in Educational Psychology Review by Kiran Bisra and colleagues at Simon Fraser University combined 64 studies with nearly 6,000 participants. Prompting learners to explain material to themselves produced an average effect of g = 0.55. It even outperformed being handed an explanation by an instructor, which suggests part of the benefit comes from generating the explanation yourself. Open-ended prompts worked better than picking an explanation from a multiple-choice list.
You can do this alone, or with a study partner where you take turns explaining. The second version brings in the social side that makes working well with others pay off.
Give it a go: at the end of each page, ask “why is this true?” and “how does this connect to the last page?” Answer in full sentences, not keywords.
8. Put your phone in another room
You know phones are distracting. The question is how much.
Quan Chen, Zheng Yan and colleagues at the University at Albany answered that in a meta-analysis published in Computers in Human Behavior in September 2024. They combined 27 randomised experiments with 2,245 participants averaging around 21 years old. Texting, chatting, email and social media during learning had a medium-sized negative effect on immediate recall, and a large one for lecture material specifically. For reading, the evidence was too thin to call, with only four studies.
A likely reason: every glance at a notification breaks the thread you were building, and you pay a re-entry cost each time you come back. Deep focus, the kind people describe as being in the flow state, needs long stretches without that interruption. Before you sit down, leave the phone charging in a different room, and set a timer on something else.
9. Make AI quiz you instead of answering for you
AI chatbots can make you feel like a genius during practice and leave you stranded on test day.
The clearest evidence comes from a field experiment by Hamsa Bastani and colleagues at the Wharton School, published in PNAS in June 2025. Nearly 1,000 high school maths students in Turkey practised with either a standard ChatGPT-style tool, a version designed as a tutor that gave hints instead of answers, or no AI at all. During practice, the standard tool raised scores by 48% and the tutor version by 127%. Then the AI was taken away for the exam. Students who’d used the standard tool scored 17% lower than those who never had it, while the tutor group did about as well as the no-AI group. The authors concluded that “unfettered access to GPT-4 can harm educational outcomes.”
Guido Makransky, an educational psychologist at the University of Copenhagen, made a similar point to Scientific American in June 2025: chatbots work best when they prompt you with reflective questions rather than hand over answers.
Test it this week: start your AI chat with “Don’t give me the answer. Ask me one question at a time and only give hints when I’m stuck.”
10. Learn from a programme that used AI the right way
What does AI look like when it’s set up as a tutor rather than an answer key? Nigeria offers one of the first rigorous answers.
A World Bank team led by Martin De Simone ran a six-week randomised trial with first-year senior secondary students, who used Microsoft Copilot (powered by GPT-4) as a virtual tutor for English. The results, published as a World Bank Policy Research Working Paper in May 2025, showed a gain of 0.23 standard deviations in English and 0.31 across the full assessment, which also covered AI knowledge and digital skills. The authors estimate this equals roughly 1.5 to 2 years of typical schooling, which would make it one of the more cost-effective learning programmes on record. Girls and students who started with higher scores gained the most. One note: a working paper isn’t peer-reviewed yet.
The part worth copying is the setup. This was a fixed, six-week programme aimed at one subject, not open-ended chatbot use whenever a student felt like it.
Here’s a simple way in: pick one topic, write a short list of what you need to learn about it, and use AI only to work through that list, one item per session.
11. Ask for feedback that tells you what to fix
Not all feedback is equal. “Good job” feels nice and teaches nothing.
Benedikt Wisniewski, Klaus Zierer and John Hattie’s January 2020 meta-analysis in Frontiers in Psychology is the largest look at this question: 435 studies, 994 effect sizes and more than 61,000 participants. Feedback overall had a medium effect (d = 0.48), but the type mattered a lot.
| Type of feedback | Effect size (d) |
|---|---|
| High-information feedback (what was wrong, why, and how to improve) | 0.99 |
| Corrective feedback (right or wrong) | 0.46 |
| Rewards or punishment | 0.24 |
The richer the information, the bigger the payoff. A grade tells you where you stand. Specific comments tell you where to go. If asking for critique makes you squirm, know that a lot of high achievers feel the same, and plenty of them wrestle with imposter syndrome too.
Your move: after your next assignment or practice test, ask your teacher, tutor or colleague one question: “What’s the single biggest thing I should change?”
12. Drop the “learning style” label
Visual learner. Auditory learner. Hands-on learner. Plenty of study guides still devote whole sections to figuring out which one you are. The evidence says it’s not worth the time.
Virginia Clinton-Lisell and Christine Litzinger at the University of North Dakota ran a meta-analysis of the “matching” idea, published in Frontiers in Psychology in July 2024. Across 21 studies and 1,712 participants, teaching to someone’s preferred style gave a small benefit (g = 0.31). But only 26% of the outcomes showed the pattern the theory predicts, where each style learns best with its own method. The authors called the benefits “too small and too infrequent to warrant widespread adoption.”
The better question isn’t “what kind of learner am I?” It’s “what does this material need?” Anatomy needs diagrams. Pronunciation needs audio. Woodworking needs your hands. So when you study something new, use whichever format suits the content, and combine words with a picture whenever you can.
13. Write notes by hand when you need to think
Typing is faster. That’s sometimes the problem, because speed lets you transcribe without processing.
Here, the expert debate is more useful than any single headline. In January 2024, Van der Weel and Van der Meer published an EEG study in Frontiers in Psychology showing that handwriting, but not typing, was linked to wider connectivity across the brain in university students. It got plenty of press. A year later, in January 2025, Svetlana Pinet of the Basque Center on Cognition, Brain and Language and Marieke Longcamp of Aix-Marseille University and CNRS published a commentary in the same journal pushing back. The study hadn’t measured learning at all, they noted, and the typing condition used one finger with no on-screen feedback, which isn’t how anyone types. Even so, both researchers still endorsed handwriting’s value and called for better studies.
So take it as a reasonable bet rather than settled science. Handwriting forces you to compress and choose, and that choosing is a form of thinking.
A quick experiment: take notes by hand for one lecture or chapter this week, using only half a page. See how much you can reconstruct the next day.
14. Treat sleep as part of studying
All-nighters feel heroic. The data is less impressed.
In a 2019 study in npj Science of Learning, Okano and colleagues gave wearable trackers to 88 students in an introductory chemistry class for a full semester. Sleep duration, quality and consistency together accounted for about 24% of the variation in their grades. Later bedtimes and irregular sleep went with lower scores. It’s correlational, from a single course, so it can’t prove sleep caused the grades. A newer cross-sectional study, by Joshua Ampofo, Binghai Sun and colleagues in Frontiers in Sleep (May 2025), points the same way: among 400 university students in Tokyo and London, worse self-reported sleep quality went with weaker scores on memory, attention and reasoning tests.
During sleep, the brain replays and stabilises what you learned that day. Skip it, and you’re asking yesterday’s material to survive without its overnight save. If you’re dragging through the day anyway, these tips on fighting sleepiness at work can help in the short term.
Start small: set a fixed wake-up time for the next two weeks, weekends included, and let your bedtime settle around it.
15. Move your body before you study
A short burst of exercise isn’t a detour from studying. It may be a warm-up for it.
The broadest summary yet came out in Psychological Bulletin in February 2025. Chang, Etnier and colleagues reviewed 30 meta-analyses covering 383 studies and 18,347 participants. A single bout of exercise improved cognitive performance with a small-to-moderate effect overall (0.33), with gains for memory (0.23), attention (0.37) and executive function (0.36). You can read the abstract of the 2025 meta-review of acute exercise and cognition on PubMed. Notably, age, intensity, type and duration didn’t change the effect much, which means you don’t need a gym session.
Next time, do this: walk briskly for ten minutes, or climb a few flights of stairs, right before a study block.
16. Take breaks on a timer, not when you feel like it
Fatigue sneaks up in a predictable way. You start fresh, so you skip the break. An hour in, your attention is fraying, but you’re “nearly done,” so you push on. By the time you finally stop, you’re drained and the last stretch didn’t stick. The decision to rest gets made by the tired version of you, which is the version least able to make it well.
Fixed breaks remove that decision. A 2023 study by Biwer and colleagues at Maastricht University, published in the British Journal of Educational Psychology, tested this with 87 university students during a real day of self-study. Students who chose their own breaks studied longer and took longer breaks, but reported more fatigue and distraction, and less concentration and motivation, than those on a fixed schedule (24 minutes on, 6 off, or 12 on, 3 off). They got the same amount of work done. It’s one study, one day, with a modest sample, so treat it as a promising pointer rather than proof.
Breaks also cut down the small mid-session choices that pile into decision fatigue. Set a 25-minute timer, stand up and move when it rings, then go again.
17. Take the training wheels off early
Training wheels make riding a bike easy. They also stop you learning balance, which is the whole point. Study aids work the same way: notes open beside you, hints on every question, answers on the back of the card in view.
The analogy only goes so far. Unlike bikes, beginners do need some support. But the research points the same direction. In his 2025 review, Karpicke notes that retrieval practice is easy to undermine by giving too much support, and cites work by Agarwal and colleagues (2008) in which open-book quizzes led to more forgetting over a week than closed-book ones. He also describes a 2006 study by Carpenter and DeLosh where free recall felt harder than recognising answers (69% initial success versus 89%) but led to the best final scores. Easier during practice, worse on the day.
Feeling fluent and being able to perform aren’t the same thing. If you can only answer with your notes open, you don’t know it yet. Close the notes, attempt the problem, and only check them after you’ve committed to an answer. That’s also a fast way to improve your memory right now.
Test it this week: turn one set of open-book exercises into closed-book ones and compare how much you remember on Friday.
18. Anchor your studying to a fixed daily slot
Motivation comes and goes. Habits stay.
A December 2024 systematic review and meta-analysis in the journal Healthcare by Ben Singh and colleagues at the University of South Australia looked at how long it takes health habits to form, drawing on more than 2,600 participants. The median was 59 to 66 days, with a huge range of 4 to 335 days. What sped things up: how often people repeated the behaviour, doing it at a consistent time (morning routines helped), enjoying it, and planning ahead. Those studies were about health behaviours, not studying, but the mechanism of repeating something in the same context until it runs on cue is the same one you’re trying to tap.
So expect about two months, not three weeks, before studying feels automatic. Building it alongside other good habits to have in life helps, since stacked routines prop each other up.
Give it a go: pick one daily slot, like right after breakfast or right after dinner, and put a 15-minute study block there every day for the next eight weeks.
What to do when you slip
You’ll skip days. Everyone does. What matters is what happens next.
The reassuring part comes from a 2010 study by Lally and colleagues in the European Journal of Social Psychology, which tracked 96 volunteers building a new daily habit. Missing a single opportunity “did not materially affect the habit formation process.” One lost day is noise, not failure. The trouble starts when one missed day turns into “well, I’ve blown it” and then a missed week.
How you talk to yourself after a setback matters too. In a 2012 series of experiments, Breines and Chen gave students a deliberately hard test. Those who then heard a brief self-compassionate message (that finding the test hard is common, and not to be too hard on themselves) were more motivated to study and improve afterward than those who didn’t. Beating yourself up feels like accountability, but it tends to push people away from the task.
So when you slip, keep it boring:
- Restart at the next scheduled slot, not “next Monday.”
- Make the restart session smaller than usual, even ten minutes.
- Notice the reason without spiralling. If your mind starts looping on the failure, these ideas on stopping ruminating thoughts help.
- If the whole day went sideways, use a few of these steps to turn a bad day around and try again tomorrow.
Conclusion
Effective learning is less about willpower and more about method. Pull information out of your head instead of pouring it back in. Space it, mix it, explain it, sleep on it. Use AI as a coach that asks questions, not a ghostwriter. Then make the whole thing a routine so it runs without heroic effort.
You don’t need to overhaul everything. Choose two strategies from this list, the ones that feel most doable, and run them for the next month. Then add another.
If you want more practical, evidence-based guides on study and skill-building, have a look at the Education section of the blog.
