Practise the thinking. Keep the judgement.
JLPT N5 mock-test review: turn one attempt into three useful tasks
Review vocabulary, sentence meaning and listening separately, then choose a small follow-up task instead of immediately repeating the same mock.
Keep the conditions of the attempt
Record whether you used a timer, notes, pauses or repeated audio. These conditions affect what the result tells you. A score from an assisted review should not be treated as equivalent to your first independent attempt.
Separate correct answers you could explain from guesses. A lucky choice can identify a point worth reviewing even when the total looks comfortable. A wrong answer is useful when you can name what to practise next.
Sort a mistake into a useful task
For an unreadable word, practise its reading in a short sentence. For a sentence you misinterpreted, identify the verb and relationship. For audio, work out whether you missed the sound, expression or final decision.
An original example is 駅で友だちを待ちます. If you read 駅 correctly but confuse who is waiting, rereading a kanji list will not directly fix the error. Draw the people and action, then return to the sentence.
Retest with changed material
Change a word or situation and answer without reopening the old explanation. When you simply repeat the same mock immediately, memory of the answer can make the result look better than your transferable understanding.
Keep one changed example for each recurring error. You can retest those examples after a gap, then use a new mixed set to see whether the distinction survives a less predictable context.
Read a practice result as practice evidence
The official JLPT uses its own scoring and pass criteria. A private practice percentage or a few correct answers should not be converted into an official scaled result or a guaranteed pass.
End the review with three tasks you can actually do: one vocabulary check, one sentence contrast and one listening decision. Choose their size from the time you have, then judge the next attempt under comparable conditions.
Try it yourself
You got an item right but chose between two answers at random. Should you record it as securely understood?
Show the answer and explanation
No. Mark it as correct but uncertain and retest the distinction.
The answer count and your explanation provide different information. Preserving both gives you a better next practice decision.