You understood the whole video. Now write it down.

When someone explains a topic, you understand it. That feels the same as knowing it. But it is not the same, and board marks are lost in the gap between the two. This page shows what the research found, what works better, and how the same hours of study can get you many more marks.

Eight minutes to read, eight things to try

Feels like learning. Is learning.

Nobody chooses the left column because they are lazy. They choose it because it is comfortable. Nothing in it can go wrong. In the right column, things can go wrong. That is exactly why it is useful.

Feels like learning

  • Watching the solution and nodding along
  • Reading the chapter again
  • Highlighting the important lines
  • Making neat notes
  • Feeling confident about a topic

Is learning

  • Writing the solution with the book closed
  • Answering from memory, then checking
  • Getting it marked step by step
  • Coming back to it a week later
  • Knowing exactly which step loses the mark

That is the main idea. The rest of this page gives the evidence, and what it changes for you.

Reading again feels like learning. Writing from memory is learning.

How much you still remember after one reading, compared with five short revisions from memory spread over two months.

Research: Roediger and Karpicke, 2006, and Cepeda and colleagues, 2008, both in Psychological Science.

0255075100010203040506090% after five short revisionsnear 0% after one reading
Five spaced revisions (dots show the revision days)Read once
These curves are a model of the pattern the studies found. Each correct revision makes the next fall slower. Days are along the bottom, and percent remembered is up the side. Move across the chart to see the values.
Show as a table
DayRead onceSpaced revisions
0100%100%
185%100%
451%100%
731%88%
1019%100%
213%100%
301%91%
400%100%
600%90%

In the 2006 study, one group of students read a passage and then wrote down what they remembered. Another group read it twice. A week later, the first group remembered far more. The group that read twice felt more confident, but they were wrong. The 2008 study asked how far apart revisions should be. The answer was about 10 to 20 percent of the time left before the exam. With 90 days left, that means revising a topic every one to two weeks now, and every few days near the end.

So, on this site

Every study block ends with you writing answers from memory. The site records your marks, not how many pages you read. Topics come back for revision on a fixed schedule, so you never have to remember when to revise them.

Of ten common study methods, only two work well.

Ratings from the biggest review of study methods. It covered hundreds of experiments.

Research: Dunlosky, Rawson, Marsh, Nathan and Willingham, 2013, a review of hundreds of studies.

High usefulnessPractice testingPractice testing: high usefulnessSpaced practiceSpaced practice: high usefulnessModerate usefulnessAsking whyAsking why: moderate usefulnessSelf-explanationSelf-explanation: moderate usefulnessMixed practiceMixed practice: moderate usefulnessLow usefulnessSummarisingSummarising: low usefulnessHighlightingHighlighting: low usefulnessKeyword mnemonicsKeyword mnemonics: low usefulnessPicturing the textPicturing the text: low usefulnessReading againReading again: low usefulness
The review rated each method on how reliably it helped learning, across different students, subjects and tests. The length of each bar shows its rating.
Show as a table
MethodRating
Practice testingHigh
Spaced practiceHigh
Asking whyModerate
Self-explanationModerate
Mixed practiceModerate
SummarisingLow
HighlightingLow
Keyword mnemonicsLow
Picturing the textLow
Reading againLow

The whole plan is built on the two methods at the top. The five at the bottom are the ones most students use. That is your chance. Spend the same hours, but on the methods at the top of this chart, not the bottom.

So, on this site

You read the chapter only to fix an answer you got wrong. Reading is not the main work. Notes are for things you need to look up. They are not proof that you read the chapter.

Three correct revisions on different days make it stay.

Students got the material right from memory three times. Then they studied it again in later sessions. Months later, they still remembered it.

Research: Rawson and Dunlosky, 2011, Journal of Experimental Psychology.

One correct answer from memory tells you that you know it today. Three correct answers on different days, each after you have forgotten a little, tell you that you will know it in February. The study found that going beyond three added very little. The gain came from the gaps between the days, not from repeating more times.

So, on this site

On your mastery map, a derivation is marked solid only after you write it correctly from memory on three different days. Understanding it once is the start, not the end.

Mixed practice feels harder and gets better marks.

Students who mixed problem types while practising did worse that day. But they did much better in the test than students who practised one type at a time.

Research: Rohrer and Taylor, 2007, Instructional Science.

Textbook exercises give you twenty questions of the same type, one after another. After the third question, you stop choosing which method to use. You just repeat it. The exam does not tell you which chapter a question comes from. So choosing the method is the real skill. Practising one type at a time never trains it.

So, on this site

Practice sets mix chapters, and each day mixes two subjects. If a set feels harder than the textbook exercise, that is on purpose. It shows the method is working.

New topic: solved example first. Known topic: questions first.

Studying a solved example before trying problems helps a beginner. The same help slows down a student who already knows the method.

Research: Sweller, 1988, and Kalyuga and colleagues, 2003.

This is why a video can be the right choice on Monday and a waste of time on Friday. Watching a solution is the same as studying a solved example. It helps while the method is new. Once you can solve it yourself, watching it only costs you time.

So, on this site

For a weak topic, you start with a model answer, explained one step at a time. For a strong topic, you go straight to timed questions, and the site does not suggest a video.

Feedback that tells you what to do next is one of the strongest effects ever measured.

Not just a score. Marks for each step, the reason each mark was lost, and what to do next.

Research: Hattie and Timperley, 2007, a synthesis of 196 studies, and Bloom, 1984.

Bloom's tutors did two things a classroom cannot. They checked each topic before moving on. And they gave feedback on the same day. An automatic marker and planner can do both, for every answer, at no extra cost. That is the job of this site. Every answer you give sends back one of four signals, and each signal has its own fix.

Fast
Slow
Right
AutomaticNext revision after a long gap.
ShakyRevise again soon, with a solved example.
Wrong
Wrong ideaExplain the difference, test again today.
GapTeach from the start, with an example first.

In written answers, every lost mark gets one of five tags.

conceptprocedureslippresentationtime

Each week, the site counts your lost marks by tag. This shows what kind of mistake you are making now. The plan then works on that kind of mistake, not the whole chapter.

You cannot fully trust your own sense of what you know.

After reading again, students judge their own knowledge badly. They judge it better when they must first commit to an answer.

Research: Dunlosky and Metcalfe, 2009, and the 2006 Roediger and Karpicke study above.

This is not a fault in you. It is how memory works. When something feels easy to follow, that feeling comes from having seen it before. Reading again and watching again build exactly that feeling. The only honest test of knowing something is writing the answer yourself.

So, on this site

Before each answer, the site asks "sure or unsure". An answer you were sure of but got wrong tells us the most. That question is tested again first.

Your brain saves memories while you sleep.

During sleep, the brain stores what you learnt that day. Teenagers need eight to nine hours of sleep. An hour cut from sleep costs you more than it gives you.

Research: Walker, 2017, and Carskadon, 2011, on adolescent sleep.

What you practised in the evening is saved into long-term memory during the night. If you cut your sleep short, some of that day's work is simply not saved. Studying late into the night feels brave, but it gives worse results in tests.

So, on this site

The plan never takes time from your sleep. If a day goes badly, the evening study gets shorter, not your sleep. Breaks are part of the plan: five minutes after each 25-minute block, and twenty minutes after four blocks.

The average student tutored one to one, with mastery checked before moving on, did better than 98 percent of students taught in a conventional classroom.

Benjamin Bloom, "The 2 Sigma Problem", Educational Researcher, 1984

Why this matters even more in ISC

Three facts about the ISC paper make these findings very useful here.

Much of the paper can be predicted.In twenty years of Magnetism questions, the same six derivations and about twenty definitions keep coming back. When the list is known, practising it from memory turns straight into marks.
Marks are given for each step.Formula, substitution, unit, direction: the examiner gives a mark for each one, or does not. Good presentation is a habit, and a habit can be practised.
Time is fixed.You get about two and a half minutes for each mark. Writing on paper against the clock is the only way to practise the real exam.

The site measures exactly these three things. Can you write the known list from memory? Could the examiner give you the mark? Did you finish in time?

A plan that knows how you are doing.

A printed timetable follows the syllabus, but it cannot see how you are doing. This plan also covers the whole syllabus. But every morning it changes, based on what you actually did. It makes you a few promises.

  • Each morning it asks you one question: how many hours do you have today? The day's work is fitted to your answer, not the other way round.
  • It tells you once, in one line, what an extra hour would get you. Then it does not bring it up again.
  • It never takes time from your sleep. Breaks are part of the plan, not a failure.
  • If you miss a day, you do not get double work the next day. The plan just starts again from where you are.
  • Nothing here checks whether you studied. It only looks at the marks in your answers.

Ninety days, four stages

Day 1306090
Getting started, 5 daysLearn, 45Practise papers, 25Final days, 15
Getting startedStudying starts on day one. Each new unit starts with 3 quick questions, answered by tapping.
LearnEvery unit is studied and revised until you can do it. One timed unit test a week.
Practise papersFull papers, marked on the same day. The other study time goes to the mistakes those papers show.
Final daysNothing new. Revision from memory and full papers. Enough sleep every night.
The stage dates move with your progress, earlier or later.

The target is 100.

It stays 100 until the last day. On each paper, about four marks in every five are predictable. You get those through revision from memory, spread over time. The rest test whether you can handle a question that looks new. That is also a skill, and you practise it in the last five weeks. Your dashboard shows your predicted score, how far it is from 100, and what will close the gap. So every week the target is a clear number, not just a hope. If 100 starts to look out of reach, the method stays the same. It keeps working for the next mark.

How a week works

  1. Say how many hours you have today.One number each morning. The day's work is fitted to it. An extra hour moves roughly one more topic to solid that week. Whether you add it is your choice.
  2. Work first on the topic that gives the most marks.The planner ranks topics by how many marks each would add per hour. Revisions that are due come first.
  3. Write on paper, against the clock.Two or three questions. Take a photo when you finish.
  4. Get it marked the same day.Marks for each step, a tag for each lost mark, one next step, and what has improved since last time.
  5. Friday report.First, how much you improved since last Friday. Then your mistake tags. Then the ranking for next week.

Test it on yourself.

Near the top of this page you saw two lists. You read one list twice. You read the other once and then wrote it down. Do not scroll up. Write every word you remember from each list.

Why this site exists

A parent built this site. They were tired of timetables that could not see how the student was doing. A printed plan follows the syllabus. It does not know what you already know, or what kind of mistake cost you marks this week. This plan follows your results. Every answer you write changes what comes next.

So do not just trust this page. Try it for one week: seven days of writing answers and getting them marked. If your marks have not improved by Friday, tell us, and the plan will change. We are not asking you to believe this. We are asking you to test it.

Sources

  • Test-enhanced learning: taking memory tests improves long-term retentionRoediger and Karpicke, Psychological Science
    2006
  • Spacing effects in learning: a temporal ridgeline of optimal retentionCepeda, Vul, Rohrer, Wixted and Pashler, Psychological Science
    2008
  • Improving students' learning with effective learning techniquesDunlosky, Rawson, Marsh, Nathan and Willingham, Psychological Science in the Public Interest
    2013
  • Optimizing schedules of retrieval practice for durable and efficient learningRawson and Dunlosky, Journal of Experimental Psychology: General
    2011
  • The shuffling of mathematics problems improves learningRohrer and Taylor, Instructional Science
    2007
  • Cognitive load during problem solvingSweller, Cognitive Science
    1988
  • The expertise reversal effectKalyuga, Ayres, Chandler and Sweller, Educational Psychologist
    2003
  • The 2 sigma problemBloom, Educational Researcher
    1984
  • The power of feedbackHattie and Timperley, Review of Educational Research
    2007
  • MetacognitionDunlosky and Metcalfe, Sage
    2009
  • Why We SleepWalker, Scribner
    2017
  • The cumulative cost of additional wakefulnessVan Dongen, Maislin, Mullington and Dinges, Sleep
    2003
  • Sleep in adolescents: the perfect stormCarskadon, Pediatric Clinics of North America
    2011