How to Analyse PYQs: Topic Weightage and Error Log

Analyse previous-year questions by attempting a complete paper, checking the matching official key, tagging each question by topic and recording why you missed it. Compare topic counts across complete papers before deciding what to revise.

This is an editorial study method, not a prediction of future questions. Start with the UPSC Prelims official paper directory; the same record-keeping method can be adapted to other exams once their papers and final keys are available.

1. Match the paper to the correct key

Record the exam, year, stage, paper name, shift where applicable, and booklet series before solving. A question number alone is not a reliable identifier across booklet series. Keep the original PDF beside your notes so that diagrams, statements and answer options can be checked.

Use the conducting body's final key for marking. If you have only a provisional or coaching key, label the score provisional. Do not silently replace an official answer with an explanation that sounds convincing. Leave disputed or withdrawn questions visibly separate.

2. Attempt before reading explanations

Save your first answer and confidence before opening the key. Choose a timed attempt if you are testing exam readiness, or an untimed attempt if you are learning unfamiliar concepts. Keep those results separate: an untimed score does not measure time management.

For each wrong answer, write the cause: missing knowledge, misreading a qualifier, calculation error, weak elimination or running out of time. A correct guess also deserves a note. Correctness without the reason can hide a gap that appears again in another paper.

3. Use a consistent topic taxonomy

Assign one primary topic to every question. Keep secondary tags for questions that connect subjects, but count only the primary tag in your headline distribution. Otherwise one question can inflate several totals. Keep an unclassified bucket rather than forcing an uncertain tag.

FieldWhat to record
IdentityExam, year, paper, shift, booklet and question number
TopicPrimary subject, chapter and optional secondary tags
AttemptYour answer, confidence, time and whether attempted
EvidenceOfficial paper URL, final key URL and key status
ReviewCorrectness, error cause, concept to revise and reattempt result

4. Calculate weightage with a visible denominator

Question weightage = questions assigned to a topic ÷ all classified questions in the selected complete papers × 100. If marks differ by question, calculate marks weightage separately using marks rather than question counts. Always state the years and papers included.

For an illustrative dataset only, 18 topic questions out of 100 classified questions would be 18%. That is not a measured UPSC statistic. A count based on a handful of uploaded questions cannot represent a complete paper. Show missing questions and excluded records next to any published percentage.

5. Compare years without promising repeats

Put years in columns and topics in rows. Compare both the topic count and the share of the paper. A syllabus or paper-format change can make older years less comparable; note that boundary rather than presenting every year as equivalent.

A recurring concept is different from an identical repeated question. Track them separately. Historical frequency describes the selected papers; it does not establish the probability of the next exam asking that topic.

6. Turn errors into a revision queue

Choose the next revision task from the overlap between recurring concepts and your own mistakes. Revisit the underlying rule or concept, then reattempt the question without looking at the answer. Keep the original result so that a remembered option does not look like a first-attempt improvement.

For calculation or comprehension mistakes, write a short process correction. For knowledge gaps, link the note to the relevant syllabus heading. Review unresolved and low-confidence answers before adding more papers to the log.

Common questions

How many years should I analyse?

Use a complete, clearly defined set that fits the current syllabus and available official evidence. This method does not prescribe a universal number. State the coverage and expand it only after the current set is consistently tagged.

Can topic weightage predict the next paper?

No. It describes the historical sample you selected. Use it to organize revision alongside syllabus coverage and your own errors.

Should I combine GS and CSAT?

Keep separate paper records and separate topic tables. Combining distinct papers into one percentage can obscure the skills each paper tests.

Sources and next steps

Use UPSC's question-paper listings, the older-paper archive and answer-key section to establish document identity. Our tagging and error-log workflow is editorial guidance. Read our source and verification method before using any analysis table.