Short answer: cramming can make material feel familiar for a few hours, but familiarity is not the same as being able to retrieve and use the material later. A better emergency strategy is to reduce the scope, test yourself without notes, revisit weak material across several sessions, and protect enough sleep to think clearly. This guide does not promise a particular grade. It gives you a way to use a limited final week more deliberately. Why cramming feels more effective than it is When you reread the same page repeatedly, each pass becomes easier. That ease can be mistaken for learning because the words look familiar. An exam usually asks for something harder: recalling an idea without the page present, distinguishing similar concepts, or applying a method to a new problem. In experiments on test-enhanced learning, repeated retrieval produced stronger delayed retention than repeated study, even though repeated study could look better on a test given only minutes later. That distinction matters: a method can improve immediate fluency without producing the memory you need days later. A large review of distributed practice also found that separating learning opportunities generally supports longer retention better than concentrating them into one block. The precise spacing depends on the material and when you need it, so there is no universal “perfect interval.” The three problems to solve in the final week Coverage: You may have more material than the available time can support. Diagnosis: Rereading hides what you cannot recall. Follow-up: A topic studied once is easy to lose before exam day. The plan below addresses these in order. It deliberately leaves some material out if necessary. Honest prioritization is more useful than pretending every chapter will receive equal attention. A seven-day recovery plan Day 1: Build an exam map List every assessable unit from the syllabus, teacher guidance, and previous papers. Mark each item: A — secure: you can explain or solve it without notes; B — partial: you recognize it but cannot reliably produce it; C — missing: you cannot yet explain the core idea. Check these labels with a two-minute closed-book attempt. Do not label a topic “secure” only because the notes look familiar. Days 2–4: Run two learning loops each day Choose a small group of B and high-priority C topics. For each topic: Study one clearly defined section for 15–25 minutes. Close every source and write what you remember. Compare the attempt with the source and mark missing or incorrect parts. Answer one application question or worked example. Schedule a shorter revisit for the next day. The closed-book step is the diagnostic engine. Feeling stuck is not proof that the method failed; it shows you exactly what needs another pass. Day 5: Mix topics Exams rarely present one chapter at a time. Mix questions from several units and decide which method or principle applies before solving. Keep an error log with three columns: the question type, why the error happened, and what cue should trigger the correct approach next time. Day 6: Simulate, then repair Complete a timed set under realistic conditions. Spend at least as long reviewing the attempt as completing it. Sort errors into knowledge gaps, method errors, misread questions, and time-management problems. Repair the highest-frequency cause rather than rereading everything. Day 7: Short review, not a new marathon Use brief recall prompts, formulas, definitions, and your error log. Avoid opening an entirely new low-priority unit late at night. Prepare materials for the exam and follow the sleep guidance appropriate for you and your circumstances. A 45-minute emergency session 5 minutes: select one narrow outcome. 15 minutes: study one explanation or worked example. 10 minutes: reproduce it without looking. 10 minutes: answer a fresh question. 5 minutes: record the error and next review time. What this plan cannot do Retrieval and spacing cannot replace missing prerequisite knowledge, feedback from a teacher, accommodations, or enough total learning time. Some subjects require performance, laboratory practice, speaking, or extended writing rather than flashcards. Adapt the retrieval task to the real assessment: solve, explain, draw, compare, write, or perform. How to decide what deserves your limited time When the exam is close, the hardest decision is often not how to study but what to leave out. Use four questions for every topic: How likely is it to appear? How large is your current gap? Does it support several other topics? Can you improve it meaningfully in the time available? A high-priority topic is assessable, currently weak, connected to other material, and small enough to repair. A low-priority topic may be obscure, independent, and too large to learn safely in one week. Give each topic a simple priority of high, medium, or low after a closed-book attempt. Do not rank by anxiety alone. Students sometimes spend an entire evening on the chapter that feels most frightening even when it represents a small part of the assessment. Check the syllabus, marking scheme, teacher guidance, and representative papers before allocating time. Protect prerequisite knowledge first. If you cannot balance an equation because the underlying algebra is missing, repairing the algebra may unlock several chemistry questions. If an essay depends on three core case studies, retrieving those cases may be more useful than collecting ten additional examples you cannot explain. Adapt retrieval to the subject Retrieval does not mean using flashcards for everything. The retrieval task should resemble the action required in the assessment: Mathematics and physics: choose an unseen problem, identify the method without a label, solve it, and check every decision. History and literature: produce a claim, supporting evidence, and a limitation from memory before writing a timed paragraph. Biology and geography: draw and label a process, explain the causal links, and compare the result with the source. Languages: retrieve vocabulary inside sentences, respond aloud to prompts, and correct pronunciation or grammar with reliable feedback. Computer science: predict what code will do, write a small function without copying, test it, and explain the error. The common feature is an attempt before checking. Copying a model solution can help you understand a first example, but it does not show whether you can select and execute the method independently. A worked example: prioritizing a biology exam Imagine an exam containing cell biology, genetics, ecology, and experimental methods. A quick diagnostic shows that you can explain cells, partly understand genetics, cannot interpret several experiment questions, and have not reviewed ecology. Previous papers show that experimental methods appear across all four units. A sensible first priority is experimental methods because it is weak, frequent, and connected to many questions. Genetics may come next because the gap is partial and recoverable. Ecology needs a narrower target: choose the most assessable processes instead of trying to rebuild the entire unit in one evening. A session could begin with one experiment question completed without notes. After checking, record whether the problem was terminology, graph interpretation, variable control, or evaluation. Study only that missing element, then answer a second question with a different context. The next day, retrieve the same reasoning using a third example. This sequence produces evidence of improvement rather than a vague feeling that you reviewed the chapter. Use an error log that changes the next session An error log is useful only when it changes what you practise. For each error, record the task, the cause, the corrected principle, and the next test. Keep cause categories concrete: Knowledge gap: a fact, definition, formula, or step was missing. Selection error: you knew several methods but chose the wrong one. Execution error: the method was correct but a calculation or sequence failed. Interpretation error: you misread a command word, graph, source, or condition. Communication error: the reasoning was present but not expressed with enough evidence or precision. After three sessions, look for repetition. If the same cause appears several times, create a focused repair set. Repeated interpretation errors call for command-word practice; repeated method-selection errors call for mixed problems; repeated missing definitions call for short retrieval prompts with feedback. Plan each day without creating another marathon A recovery week still needs limits. Begin with the hardest high-priority task during your most reliable period. Use later sessions for shorter retrieval, checking, and error repair. Keep meals, travel, classes, care responsibilities, and sleep visible on the plan before adding study blocks. A timetable that ignores fixed commitments is not realistic. At the end of each day, answer four questions: What can I now produce without notes? Which error repeated? What must be revisited tomorrow? Which planned task should be removed because time is no longer available? Removing a low-value task is a planning decision, not a failure. If you miss a day, do not squeeze two full days into the next one. Recalculate the remaining sessions, protect the highest-priority retrieval and simulation work, and reduce the scope. The plan exists to support decisions; it should change when the available time changes. Frequently asked questions about last-week revision Should I stop making notes completely? No. Brief notes can organize a confusing explanation or record a correction. The problem is using transcription as the final activity. After making the note, close the source and use it to answer, explain, solve, compare, or draw from memory. Are flashcards useful in the final week? They can help with compact knowledge such as terms, formulas, vocabulary, and process steps. They are less suitable as the only preparation for essays, unfamiliar calculations, source analysis, or performance. Match the card to a real retrieval need and include application work. What if retrieval feels slower than rereading? It usually produces more visible difficulty because it exposes gaps. That discomfort is information. Keep attempts short enough to complete, check them promptly, and use the result to choose the repair. Speed can improve after the knowledge and method become more secure. Can this plan guarantee that I pass? No. Results depend on prior knowledge, assessment difficulty, available time, health, support, and many other conditions. The plan improves the quality of the decisions and practice you make with the remaining time; it cannot manufacture missing weeks of learning or guarantee an outcome.