How to read a mock test result
The score is the least useful number on the result page. Here is what to pull out of a mock instead, and roughly how long it should take.
Most people read a mock result for about ninety seconds. Score, rank, a quick look at which section went badly, then on to the next test. That is a waste. The paper you just sat is the most specific information you have about your own preparation, and the score is the part of it that tells you the least.
Here is a routine worth trying instead. It takes about an hour and you only need a notebook.
The score is not comparable to anything
Paper difficulty swings hard between mocks, even inside one institute's series. A 96 on one paper and an 82 on the next can be the same performance. So the absolute number means little on its own, and comparing it to a friend's score from a different mock means less.
Two numbers travel better:
- Where you sat relative to everyone who took the same mock.
- Your accuracy on the questions you chose to attempt.
Rank within the same paper handles difficulty for you. Everyone in that cohort got the same brutal set of questions. If you dropped twenty marks but held your position, the paper got harder. Nothing about you changed.
Accuracy is the number to watch
Count four things: attempted, correct, incorrect, left blank. Then work out correct ÷ attempted.
In Prelims a wrong answer costs a third of the marks for that question, so every attempt is a bet with a fixed price. If your accuracy on attempted questions sits around 60%, more reading will not fix your score in the near term. Attempt discipline will. You are paying for guesses that do not cover their own cost.
Sort the wrong answers into three piles
Every incorrect answer is one of three things. The fix is different for each, which is why mixing them together lets people revise for weeks without improving.
- You did not know it. Content gap. Note the topic and go read it.
- You knew it and still got it wrong. Misread the question, missed an "except", grabbed the first plausible option, ran out of clock. This is not a content problem, and reading more will not touch it.
- You narrowed it to two and picked wrong. Partial knowledge. Usually the quickest thing on this list to fix, because you are one distinction away.
Go through your wrong answers and write a 1, 2 or 3 next to each. Twenty minutes, no deep analysis. Then count the piles.
Most people find pile 2 is bigger than they expected, sometimes bigger than pile 1. If that is you, your problem this month is how you read questions under a clock, not what you know. No amount of fresh material fixes it.
Mark the ones you got right by luck
Do this while the paper is fresh, because you will forget by evening. Any question where you guessed outright, or narrowed to two and got lucky, put a tick beside it.
Those marks are on loan. If you had four lucky hits, your real score is four right answers lower, and next time the coin lands the other way. This is also the fastest way to catch a score that is flattering you.
Where the time went
If your test platform records per-question timing, find every question you spent more than two minutes on. Two patterns show up.
Long time, wrong answer: you should have left it. That is a marking decision, not a knowledge gap.
Long time, right answer, on a topic you know well: you did not trust yourself. That hesitation costs you the questions at the end you never reached.
No timing data? Write down section-level time from memory as soon as you finish. Rough numbers are fine.
Do not over-read a single paper
One mock gives you maybe two or three questions per topic. That is not enough to conclude you are weak in Polity. It is enough to conclude that you misread three questions, or that you attempt too many, because those patterns show up across the whole paper rather than inside one topic.
Topic-level judgements need five or six mocks before they mean anything. So keep one line per mock:
| Date | Attempted | Accuracy | Most common error |
|---|---|---|---|
| 12 Aug | 94 | 61% | misread the question |
| 19 Aug | 88 | 68% | misread the question |
| 26 Aug | 82 | 74% | narrowed to two |
Three lines in and the story is clearer than any single score was: fewer attempts, better accuracy, one error type that has not shifted.
Then change exactly one thing
Pick one change for the next mock, and make it something you can check afterwards. "No more than 90 attempts." "Leave anything that runs past two minutes."
One change, one mock, then look at the number. Two changes at once and you will not know which one worked.
That is the method. It is unglamorous and it eats an hour you would rather spend on new material. It is still the best hour in your week.
A note on why we are writing this: Exameow is being built to do the tagging and the across-mocks trend work for you, off your own attempt history. It is not live yet. Until it is, the notebook version works fine.