Science is the subject where the gap between "I studied it" and "I can answer it" is widest.
You read the chapter. You understood the chapter. Then the paper asks you to label a diagram you have only ever looked at, balance an equation you have only ever read, and derive a number from three given values in a sentence that hides which formula applies. None of those is a reading task, and reading is what most students did.
This post is about the specific things a Science paper rewards, in the order they are worth fixing.
Science Is Four Subjects Wearing One Cover
Physics, Chemistry and Biology behave completely differently under revision, and treating them as one block is the first mistake. A fourth thing - the practical and the diagram work - runs across all three and is where the cheapest marks in the whole paper sit.
Physics is a problem-solving subject. It rewards working problems, not reading theory. You cannot read your way to competence in numericals, and the number of students who try is enormous.
Chemistry is part memory, part method. Reactions, valencies and the periodic table are memory. Balancing and calculating are method. The two need different revision and get lumped together.
Biology is mostly recall and diagrams. It is the most memorisable of the three and therefore the one where good technique pays off fastest.
The practical and diagram content is everybody's easiest marks and everybody's biggest waste. A labelled diagram is worth the same as a hard numerical and takes a tenth of the time to become reliable at.
If your revision plan says "Science: 2 hours", you do not have a plan. It should say what you are doing in which of the four.
Start By Finding Out What You Cannot Do
Before any revision, one hour, honestly spent, saves ten.
Take a past paper. Sit it timed, closed-book, with no help. Then mark it strictly - not generously, strictly, the way a marker who does not know you would. What comes out is a list of specific failures, and that list is your revision plan for the next month.
Most students skip this because it is uncomfortable and because a bad score early feels discouraging. It is the opposite: a bad score in July is information, and a bad score in the hall is a result.
Sort every lost mark into one of four causes. This is the single most useful thing in this post:
- I did not know the content. Revise the topic.
- I knew it and could not apply it. Practise that question type. Reading the theory again will not help.
- I made an arithmetic or unit error. Slow down and check. This is not a knowledge problem and treating it as one wastes weeks.
- I ran out of time. A timing problem, fixed by practising to the clock, not by learning more.
Students routinely respond to cause 3 by re-reading theory, which fixes nothing and feels productive. Separating the four is what makes revision efficient, and it is a straightforward act of metacognition - thinking about your own thinking rather than just doing more of it.
Keep one page titled "my actual errors". Every time you get something wrong, write the question type and which of the four causes it was. By the third week the page has a pattern, and that pattern is worth more than any guide book's chapter summary.
Physics: Work Problems, Do Not Read Them
The failure mode in Physics is specific and almost universal.
You read a worked example. You follow every line. You feel you understand. Then a similar question defeats you, and you conclude you need to read the theory again.
You do not. Following someone else's solution and producing your own are different skills, and only the second is examined. This is the generation effect working against you: producing an answer builds far more than reading one, and reading one produces a confident feeling of competence that is not evidence of anything.
The routine that actually works
Cover the solution before you start. If the worked answer is visible, you will read it. Cover it with paper and attempt the problem first. Every time.
Write down what you are given and what is asked. In Physics most difficulty is not in the algebra but in deciding which relationship applies. Listing the given quantities with their units makes the choice visible, because the units narrow the options.
Do the units as arithmetic. Dimensional analysis is the cheapest error check available: if your answer comes out in metres per second and the question asked for a force, you made a mistake, and you know it before the marker does. Carry the units through every line rather than adding them at the end.
Learn the derivations that are asked for, by deriving them. A derivation copied three times is memorised badly. A derivation you can rebuild from the starting relationship is memorised properly, and it survives a rephrased question.
Do three of the same type, then stop. One solved problem is not a learned method. Three unaided is. Four is usually waste - move to the next type.
Numericals are a separate skill from theory
Split your Physics revision explicitly. Theory sessions: definitions, laws, statements, derivations. Numerical sessions: nothing but problems, with a pen and paper and the solutions covered.
If you cannot say which of the two you did yesterday, you probably did theory - it is the more comfortable one, and it is not what the numericals section tests.
On scientific notation and significant figures: these lose marks quietly all year. An answer of 0.000042 written as 4.2 x 10^-5 is the same number and one of them is what the scheme expects. Find out what your paper expects and be consistent.
Chemistry: Split Memory From Method
Chemistry punishes students who revise it as one activity, because it is genuinely two.
The memory half
Valencies, formulae, common reactions, the reactivity series, the properties that come up every year. These are discrete facts and they respond extremely well to spaced repetition - short daily review of a card deck rather than one long session a week.
Fifteen minutes a day on a small deck of reaction cards will do more for your Chemistry mark than two hours of re-reading the chapter, and it is the least popular advice in this post because it is boring and it works.
Write the reaction, do not read it. Reading "zinc plus dilute sulphuric acid gives zinc sulphate plus hydrogen" teaches you to recognise it. Writing it from a prompt teaches you to produce it, which is what the paper asks.
The method half
Balancing a chemical equation is a procedure. So is a mole calculation, so is anything involving stoichiometry. Procedures are learned by repetition with feedback, exactly like Physics numericals, and they are learned badly by reading an example.
The specific thing that costs marks: an unbalanced equation in an otherwise correct answer. Get into the habit of counting atoms on both sides before you move on, every single time, until it is automatic. It takes ten seconds and it is worth marks on several questions.
Units and quantities. Mole calculations are where unit errors concentrate. Write the unit at every step. If the units do not cancel to what the question asked for, something is wrong upstream.
If you cannot write the balanced equation for the five or six reactions that appear every year, no amount of theory revision will fix your Chemistry mark. Those are recall items and they need recall practice, not reading.
Biology: Recall and Diagrams, Deliberately
Biology is the most improvable of the three because most of what it asks for is retrievable content, and retrieval has a technique.
Close the book and write. Read a section once, close it, write down everything you remember, then open it and check what you missed. This is active recall and it is uncomfortable precisely because it works - the effort of retrieving is the thing that builds the memory. Re-reading is comfortable and close to useless by comparison, which is the finding students find hardest to accept.
Use the terminology. Biology marks attach to specific words. "The thing that carries oxygen" and "haemoglobin" are not equally marked. When you write from memory, write the term, and if you cannot produce the term, that is the gap.
Test yourself on the process, not the label. Many questions ask what happens and in what order. Write the sequence out as a list from memory. If a step is missing, you have found the gap that would have cost you the marks.
Long-answer questions want identifiable points. A marker is looking for a number of distinct pieces. Four marks means four points, written so they can be found - not one dense paragraph containing them. Structure is worth marks independently of content.
Diagrams: The Cheapest Marks on the Paper
This section is short and it is the highest return per minute in the whole post.
Diagram questions are worth real marks, they recur every year, and they are almost entirely predictable. Yet most students have never drawn one from memory - they have only looked at them.
Make a list of every diagram your syllabus can ask for. There are not many. Go through the textbook and write them down.
Draw each one from memory, then compare. Not trace, not copy - draw with the book closed and then check. Do this three times on separate days and it is permanent.
Label everything, spelled correctly. Labels are what carries the marks. A beautiful unlabelled drawing scores poorly and a plain fully-labelled one scores well.
Practise the apparatus diagrams too. Experimental setups are diagrams like any other and they get neglected because they feel like practical work rather than revision.
In the hall, use a pencil and a ruler. Straight lines to labels, no crossing leader lines, no shading you were not asked for. Presentation matters here more than anywhere else on the paper, because the marker has to identify what you meant.
The Practical and the Experiment Questions
Marks for practical content are among the most reliably lost, because they are learned in a lab and revised nowhere.
Know the aim, the apparatus, the procedure, the observation and the conclusion for every prescribed experiment. Those five headings are the shape of almost every experiment question, and having them as a template means you always have a structure to write into.
Know why each control exists. "Why was the second test tube kept without the enzyme?" is a comprehension question about experimental design, not a recall question, and it is asked constantly. Understanding the idea of a control - the whole logic of the scientific method - answers a family of questions rather than one.
Practise reading a result and stating a conclusion. Given this observation, what does it show? That is a distinct skill and it is trainable in an afternoon with past papers.
If your school's laboratory work has been thin, which is a real situation in many schools, you can still learn the procedures on paper and you should. Watching a demonstration of a titration is better than nothing, and knowing the written procedure is what the paper actually tests.
Free Material Worth Using
Practical, for a student in Nepal with a data plan and a phone.
Your prescribed textbook first. The Curriculum Development Centre publishes the official curriculum and textbooks, and the scope of the syllabus is what defines what can be asked. An excellent resource covering the wrong scope is a waste of your evening.
Past papers, from the source. The National Examinations Board is where official examination information lives. Rumours about pattern changes circulate every year and are frequently wrong.
Simulations for the things you cannot see. PhET has free interactive simulations for physics and chemistry concepts - circuits, gas laws, waves - and they are genuinely good for building the intuition a static diagram does not.
Free explanations when the textbook does not land. Khan Academy and OpenStax are free, credible and cover the concepts. Both are written to an international syllabus, so use them to understand and go back to your prescribed book for scope and notation.
A Six-Week Science Plan
Concrete, because "revise Science" is not a plan.
Week 1 - diagnose and clear the cheapest marks. One full past paper, timed and marked strictly, sorted into the four causes. Then the diagram list: write it out and draw every one from memory once. That single week usually moves a mark more than any other.
Week 2 - Physics numericals only. Solutions covered, units carried through every line, three problems per question type. Nothing but problems. Keep the error page running.
Week 3 - Chemistry, split deliberately. Daily fifteen minutes on the reaction and valency deck. The rest of the session on balancing and calculation, with the answers covered.
Week 4 - Biology by recall. Read once, close, write, check. Terminology written out, processes written as ordered lists. Long answers practised as identifiable points rather than paragraphs.
Week 5 - practicals and the topics your error page keeps naming. By now the page names three or four things. Those are the whole agenda. Plus the five-heading template for every prescribed experiment.
Week 6 - two more full papers to time. Marked strictly. Compare the four-cause split against Week 1: if timing errors have not fallen, practise the clock rather than the content.
Then keep the Chemistry deck and the diagram set in rotation until the exam. Both decay if left, and both are cheap to maintain - the forgetting curve is steep for isolated facts and nearly flat for facts you have retrieved a few times over weeks.
Where Science Marks Actually Go Missing
The recurring failures, in the order they cost the most.
Reading theory instead of working problems. The most common and most expensive mistake in Physics.
Unlabelled or unpractised diagrams. Free marks, thrown away every year by students who could have secured them in one afternoon.
Unbalanced equations. A correct answer marked down for ten seconds of counting not done.
Unit errors and dropped powers of ten. Not a knowledge problem, and repeatedly misdiagnosed as one.
Long answers written as one block. The points are in there and the marker cannot find them, so they are not awarded.
Practical questions never revised. Learned in a lab, forgotten by March, asked in the paper.
Never having sat a paper to time. You find out about your timing in the hall, which is the worst place to learn it.
The Short Version
Science is four subjects and they need four different treatments. Physics is problems, not reading. Chemistry is a memory half and a method half, revised separately. Biology is recall, done with the book closed. Diagrams and practicals are the cheapest marks on the paper and almost nobody drills them.
Start with one honest timed paper and sort every lost mark into four causes - content, application, care, or time. Those four need four different responses, and treating a care problem as a content problem is why some students work extremely hard and improve very little.
Then cover the solutions before you attempt anything, carry your units through every line, and draw every diagram from memory at least three times. That is most of it, and it is a different month from the one most students spend re-reading chapters they already understood.