How to Revise a Subject You Are Failing: What to Cut

To revise a subject you are failing, pick a target grade, find the earliest topic in the chain that is broken, and spend your remaining hours in proportion to the marks each topic carries. You cannot revise all of it before the exam, so decide what to leave out while you still have time to act on the decision.

This page is about one specific situation: you are turning up, you are doing the work, and you are still failing, with an exam close enough that time is the binding constraint. That is a triage problem rather than an effort problem, and the four decisions below are what triage actually consists of. If the real issue is a pile of work you never handed in, start with recovering a backlog instead. If one exam went badly and the rest of the subject is fine, the job is smaller and reading that single script closely is the faster route.

Why does revising the current topic not fix a failing grade?

Because in a cumulative subject, the topic you are failing now is usually downstream of a gap that opened one or two topics earlier. Chapter seven questions are built out of chapter four. A student who responds by revising chapter seven harder is polishing the symptom while the cause sits three weeks back in the syllabus, and the extra hours return almost nothing. This is the single most expensive mistake available in a failing subject, because it feels exactly like working hard.

There is a reliable tell that separates the two cases, and it is about where an answer dies. A question you get wrong at the end is usually a current-topic problem: you recognised the task and made a start, and the method ran out before the answer did. A question you cannot start at all is usually a prerequisite problem: you do not know what the question wants you to do, which means the idea underneath it never landed. Sort your wrong answers into those two piles before you plan a single revision session.

The second pile is the one that decides your revision order, and almost nobody sorts for it. Students who are failing revise what is physically in front of them, because that is what the class is covering. The chain runs the other way: your marks on the current topic are capped by the weakest link below it.

How do you find the earliest broken link?

By working backwards from the questions you could not start, one level at a time, until you reach something you can explain without looking it up. That is the floor, and the level directly above it is where your revision begins. The whole search takes about half an hour with a marked paper and a blank page.

Two things make this search go wrong. The first is going too far back: a student who decides the real problem is year-nine algebra has bought a project rather than a revision plan. Stop at the first level that holds. The second is skipping the test step, because the feeling of understanding a familiar topic is unreliable in exactly the subject you are failing.

The chain is tighter in some subjects than others. Maths, physics, chemistry and most of economics are strongly cumulative, so the backwards search lands on a specific topic. History, literature and essay-based social science chain more loosely, and what transfers between units there is a skill. The search still works; it just terminates on something like constructing an argument from evidence instead of on a chapter number.

What should you revise when you cannot revise everything?

Whatever the paper pays the most for, which is rarely what the syllabus lists first. Marks are not spread evenly across a specification, and the ranking is published: boards state assessment weightings in the specification, and every past paper prints a mark value next to each question. Fifteen minutes with a past paper and a highlighter produces a topic-by-topic mark count most students never build.

Once you have that count, three rules turn it into a timetable. Compulsory sections outrank optional ones, because a topic in a section you are allowed to skip can be worth a lot on paper and nothing to you. A topic that appears on every past paper in the pile outranks a topic that showed up twice in six years. And a topic the chain depends on outranks its own mark value, since fixing it moves marks in every question built on top of it.

Then there is the part that matters most when you are failing, which is where inside a question the cheap marks live. The first two or three marks of a long question are usually the definition, the statement of the rule, or the setting up of the method. Those come straight off the lower levels of the chain you have just repaired, and they are available across every long question on the paper. The final marks on the same question need the whole topic to be secure. A student at 35 percent who chases the last two marks of the hardest question has picked the most expensive marks on the paper.

Are you revising for a pass or for a good grade?

Decide this before anything else, because the two targets select different material and chasing both gets you neither. A pass is usually assembled from the low-tariff marks spread across many topics, which means breadth and the opening parts of questions. A good grade needs depth in specific topics and the long questions that carry it. Those are two different revision plans, and four weeks will fund one of them.

The useful thing is that the target converts into a number. AQA states that grade boundaries “show the minimum number of marks you need for each grade, and are published on results day”, and it publishes the tables as downloadable files. Look up the most recent ones for your specification, subtract what you currently score, and the gap stops being a feeling. The same page notes that boundaries are set after marking, once senior examiners can see how hard students found the paper, so a past series is an estimate rather than a promise. It is still a better planning number than a target grade held in your head.

Be honest about what the number says. If you are sitting at a fail with a month to go, a plan aimed at the top grade is a plan to run out of time in week three with the basics still shaky. Aim at the boundary directly above you, hit it, and then reassess. A grade you actually reach is worth more than a grade you were theoretically revising for.

What should you deliberately drop?

Something is going unrevised either way, so the only real question is whether you choose it in week one or discover it the night before. Write the drop list down, so the decision stays made.

Four things are usually safe to cut, in roughly this order:

• Topics in optional sections you will not answer. If the paper lets you choose three questions from five, two topics are free. Pick which two now.
• Low-tariff topics nothing else depends on. Check the second half of that clause carefully. A small topic holding up two larger ones is not a small topic.
• The final marks on the hardest question type. Dropping the last three marks of an extended question while still answering its opening parts is often better value than dropping a whole topic, and it keeps the rest of the paper intact.
• Everything stacked on top of a break too deep to repair. If the level underneath a whole branch of the subject cannot be rebuilt in the weeks you have, the marks above it are out of reach anyway. Write off the branch, and stop paying for it.

One rule holds across all four: never drop a topic the chain still depends on, however little it is worth by itself. The whole point of the backwards search was to find those, and cutting one undoes everything stacked above it.

Which revision method earns the most from the hours you have left?

Testing yourself, because with hours this scarce the method stops being a matter of taste. Dunlosky, Rawson, Marsh, Nathan and Willingham reviewed ten common study techniques in Improving Students’ Learning With Effective Learning Techniques and rated each one for how broadly its benefits hold up. Practice testing and distributed practice came out with high utility assessments. The review is blunt about the other end: “Five techniques received a low utility assessment: summarization, highlighting, the keyword mnemonic, imagery use for text learning, and rereading”, and it recommends that “other techniques should be used in their place (e.g., practice testing instead of rereading)”.

Read that for what it is. It rates how generally each technique works across ages, materials and test types, and it promises nothing about your grade in your subject. What it does tell you is that the two habits most students fall back on when they are frightened are the two with the weakest general case.

In triage there is a second reason to test, separate from how well it builds memory. Testing measures. You cannot tell a topic you have reread three times from a topic you actually know, so a plan built on your own confidence rests on the one instrument this situation has already proved unreliable. A closed-book question settles it in ninety seconds. For the mechanics of the two high-utility techniques together, see active recall against spaced repetition.

How do you build a revision plan for a subject you are failing?

Spend the first session on the decisions rather than on the content. That session produces four things: the level your chain breaks at, a topic-by-topic mark count from a past paper, a target expressed as a mark rather than a grade, and a written drop list. Nothing else you do this month is worth as much per hour.

From there the weeks have a shape. The first block goes to the broken level, tested closed-book until three questions in a row hold. The second block works upwards through the chain in order. The third block goes to the highest-mark topics that are still standing, and the opening parts of long questions across the whole paper. Timed past-paper sections come last, and if you have never rehearsed the real thing under a clock, revising properly for mocks is where that habit gets built.

Keep one thing in reserve: re-test the earliest level briefly once a week, because that is the material everything else rests on and the material you stop looking at first. If this subject shares a timetable with three others, splitting revision across multiple exams covers how to divide the week before you divide the subject.

Where GeniusPal fits when you cannot tell what is broken

The hard part of everything above is measurement. You need to know which topics are actually broken, and self-assessment is the one tool a failing subject has already disproved. GeniusPal turns a document you upload into a set of questions, so a chapter of your own notes becomes something you can be wrong about in a way you can see. It measures, and it cannot fix a grade.

Two parts of it matter more than generating new sets. The daily review is free on every plan, runs 20 questions, and is built from the questions you got wrong, which is the cheap repeated testing this situation calls for. Redrilling the questions you got wrong is also free on every plan. Between them, a topic you got half right stays in front of you without spending anything.

On the free plan you get two study-set generations for the lifetime of the account, ten questions per set, every question multiple choice, two full runs of each set, and one exam-style analysis for the account lifetime. The Student plan is $14.99 a month for 100 generations, sets of up to 30 questions, ten exam-style analyses a month, and the same set run as flashcards and as written recall as well as a quiz. There are three study modes in total: quiz, flashcard and recall. Uploads go up to 10 MB on every plan.

None of that decides which topics to cut. That decision is yours, and it is the one thing on this page that changes the outcome, so make it in week one while you still have a month to spend on the answer.