Some subjects are mostly understanding. Social Studies is mostly remembering: dates, definitions, figures, treaties, rivers, articles of the constitution. So are large parts of Science, and the literature components of Nepali and English.
For those, "study harder" is not advice. Remembering is a skill with known techniques, and most students have never been taught any of them.
This post covers the ones that actually work, why they work, and where each one is worth the effort.
First: Why You Forget
You are not forgetting because your memory is bad. You are forgetting because that is what memory does with information it has only encountered once.
Two mechanisms explain nearly all of it.
The forgetting curve. Retention drops steeply in the first days after learning something and then levels off. A chapter read once on Sunday is largely gone by the following weekend. This is normal and it is not a defect.
Interference. Similar material competes with itself. Study two treaties with similar names back to back and you will reliably produce the right fact about the wrong one. This is why students who "know" a topic still swap details under pressure.
Both point to the same fix, and it is not more reading. It is retrieving the information repeatedly, spaced out over days, in a form that separates similar items from each other.
The Two Techniques That Matter Most
Everything else in this post is a refinement of these two. If you only adopt these, you will remember several times more than by re-reading.
Test yourself instead of reviewing
Close the book. Write down what you remember. Then check.
This is active recall, and the advantage it holds over re-reading is one of the most reliably reproduced findings in learning research - the testing effect. The reason it works is the reason it is unpleasant: pulling information out of memory is effortful, and the effort is what strengthens the trace. Reading feels smooth and leaves almost nothing behind.
The practical version: stop writing summaries, start writing questions. After each topic, write five questions and nothing else. Answer them cold three days later.
Space the reviews out
Four ten-minute reviews across two weeks beat one forty-minute session, for the same total time. This is distributed practice, and it works because a little forgetting between reviews forces a real retrieval rather than a recognition.
You do not need software. Keep your question sheets in one folder, write the date at the top each time you answer one cold, and always pick the sheet with the oldest date. That is spaced repetition with a pen. Move sheets you have got right twice to the back - your attention belongs on the ones you keep failing.
A schedule that is good enough: review a topic the same day, then after three days, then after ten, then after a month. Four short visits and it will still be there in the exam.
Chunking: Make the List Shorter
Working memory holds only a few items at once. Chunking is the trick of turning many items into fewer by grouping them, which is why a phone number is remembered in blocks rather than as ten separate digits.
For exam material this means: never memorise a flat list. Impose a structure on it first.
Twelve causes of an event is a hopeless list. The same twelve grouped into three political, four economic and five social causes is three things to remember, each containing a small set. You will also produce a better-organised answer, because the grouping is a paragraph structure.
The same applies to Science. The functions of the liver as a list of eight is hard. Grouped into storage, production and breakdown, it is three.
Do the grouping yourself rather than taking the textbook's. The act of deciding what belongs together is elaborative encoding, and it is most of why this works.
Mnemonics: Worth It for Arbitrary Sequences
A mnemonic converts something meaningless into something memorable - usually a phrase whose first letters map onto the items you need.
They are genuinely useful for a narrow category: arbitrary lists where the order matters and there is no logic to lean on. The order of colours in the spectrum, the sequence of steps in a process, the planets, the reactivity series.
They are a poor substitute for understanding. A mnemonic tells you that seven items exist in an order; it does not tell you what any of them means. Students who memorise a mnemonic for a process they do not understand can name the steps and cannot answer a question about why step three follows step two - which is what the marks are for.
Two rules make them work:
Make your own. A mnemonic you invented is far stickier than one you were given, and it can use whatever is memorable to you - Nepali words, your friends' names, something absurd. Absurd is better; ordinary is forgettable.
Use them for the sequence and learn the content separately. The mnemonic is an index, not the book.
The Method of Loci, for the Genuinely Long Lists
The method of loci - the "memory palace" - places items along a route you know well. Your walk from home to school, or the rooms of your house in order. Each item gets a location, imagined vividly and preferably strangely.
It sounds ridiculous and it demonstrably works, because spatial memory is far stronger than memory for abstract lists, and you already have the route stored for free.
Worth the setup cost for a handful of things: a long chronology in History, the sequence of an involved biological process, the articles of a document you must recall in order. Not worth it for twenty short definitions - use question sheets for those.
If you try it, be specific and physical. "The Sugauli Treaty at the front gate, with the gate itself being signed" is memorable. "Treaty near the house" is not.
Subject by Subject
Social Studies. The most memory-intensive subject and the one where these techniques pay most. Question sheets for every definition and date. Chunk anything that arrives as a list. Space the reviews. Precision matters here more than anywhere - an approximately-right date is wrong - so mark yourself strictly.
Science. Chunk the functions and processes. Draw diagrams from memory rather than copying them, because the labelling is what gets examined and drawing is retrieval. For formulas, do not memorise in isolation - work problems with them, which encodes the formula and its use together.
Nepali and English literature. Do not attempt to remember whole texts. For each one, hold three things: the author, the theme, and two or three quotable lines. That is what answers are built from. For grammar, memorising the rule is not the goal - do exercises, because the exam asks you to apply it.
Mathematics. Barely a memory subject, with two exceptions. The standard identities and formulas should be automatic, so that working memory is free for the problem rather than spent retrieving a formula. And your own error log - the running list of mistakes you keep making - is worth more than any memorised content.
Making Flashcards That Are Worth the Time
Flashcards are the standard tool for the memory-heavy subjects, and most student cards are made badly enough to waste the effort.
One fact per card. A card with five things on it cannot be marked right or wrong, so you will always half-know it and always keep it in the pile.
The front must be a question, not a topic. "Sugauli Treaty" is a topic and it invites vague recognition. "In which year was the Sugauli Treaty signed, and what did Nepal lose?" is a question with a checkable answer.
Write the answer in the shortest form that would earn the mark. Not a paragraph. If the exam wants a date and two consequences, the card holds a date and two consequences.
Make them yourself. A bought or borrowed set skips the part that does the encoding. Writing the card is roughly half of the value.
Three piles, not one. Got it right twice - to the back, review monthly. Right once - review in a week. Wrong - review tomorrow. This is spaced repetition done physically, and it puts your attention where the failures are instead of spreading it evenly.
Say the answer out loud before turning the card. Thinking "yes, I know this" and turning it over is recognition wearing the costume of recall. Committing out loud forces an actual retrieval, and it is uncomfortable exactly when you do not know it - which is the information you wanted.
Recall Versus Recognition, and Why It Matters in the Hall
One distinction explains most of the gap between how prepared students feel and how they perform.
Recognition is knowing something when you see it. It is what re-reading, highlighting and multiple-choice practice train.
Recall is producing it from nothing. It is what a written examination almost entirely requires.
Recognition arrives much earlier and feels like mastery, which is the trap. A student who has read a chapter six times has strong recognition and may have almost no recall, and there is no internal signal that distinguishes the two. The only way to find out is to try to produce the answer with the book shut.
This is also why practising with multiple-choice questions is a poor proxy for the SEE and NEB papers. Choosing between four options is a recognition task; the exam mostly asks you to generate. Use MCQs as a quick coverage check, not as your main practice.
The practical rule that follows: every study session should contain at least one moment where you produce something from an empty page. If a session never had that moment, it trained the wrong thing.
Things That Feel Like Learning and Are Not
Worth naming, because each of these consumes hours and returns very little.
Re-reading. Produces familiarity, which is easily mistaken for knowledge. You will recognise every line and be unable to produce any of it.
Highlighting. Marking a sentence as important is not knowing it. If you highlight, immediately write the question that sentence answers.
Copying notes out neatly. Transcription with better handwriting. Your hand is busy and your understanding is idle.
Listening to a recording of yourself reading notes. Passive, and it competes with nothing. Being asked the questions out loud would work; hearing the answers does not.
Studying with the book open beside you. If the answer is visible, you are not retrieving, and retrieval is the entire mechanism.
The common thread: if it is comfortable, it is probably not working. The techniques that work feel like effort, because the effort is the encoding - a pattern researchers call desirable difficulty.
Using an AI Tutor as an Examiner
The single most under-used prompt available to a student: "Ask me ten questions on the digestive system, one at a time, and tell me if I am wrong."
That turns the tool from an encyclopaedia into a retrieval machine, which is where the learning is. Almost every student uses an AI tutor to be told things; very few use it to be tested.
Three other prompts worth having:
- "Here are my notes. Turn them into twenty questions." - the conversion step, done in seconds.
- "Group these fourteen items into categories that make sense." - chunking assistance, though doing it yourself encodes better.
- "I said X. What did I get wrong and what did I miss?" - feedback on a recalled answer rather than on a read one.
One caution: verify anything factual against your textbook. A model can produce fluent, confident, wrong dates, and a date memorised from a chat window is a risk you carry into the exam.
The Week Before
Revise from your question sheets and one-page summaries, not from textbooks. If those do not exist, this is the moment you will wish you had made them - and the lesson for next term.
Do the oldest sheets first. Recency is misleading: the topics you studied last feel solid and the ones from two months ago are the ones actually at risk.
Separate the confusable pairs deliberately. Whatever two things you keep swapping - two treaties, two similar formulas, two authors - put them side by side on one page with the difference written between them. Naming the difference explicitly is the only thing that reliably stops interference.
Sleep. Memory is consolidated overnight, which means sleep is not time away from revision - it is the part of revision where the previous hours are made permanent. An all-nighter trades a few hours of poor-quality input for a measurable reduction in what you can retrieve the next morning.
The Short Version
You forget because that is what memory does with things it has met once. The fix is retrieval, repeated, spread over days.
Write questions rather than summaries. Answer them cold, oldest sheet first. Chunk any list before trying to remember it. Use mnemonics only for arbitrary sequences, and make your own. Separate the pairs you keep confusing by writing the difference down.
And distrust anything that feels comfortable. The techniques that work are the ones that feel like effort, which is the whole reason so few students use them.