Flagging questions in HKSI Paper 1, and coming back to them

Computer-based papers normally let you mark a question for review and jump back to it later. It is a small feature and candidates use it in two ways that cost them marks: as a substitute for answering, and as an excuse to revisit everything.

Used properly it does one job. It remembers which of your answers were shaky, so that you do not have to.

The two-pass model

The order matters. Do the final sweep even if you are confident you answered everything, because a mis-click during navigation is easy and invisible until it costs you a mark.

What should you flag?

Three categories, and nothing else.

• Genuine last-two questions. You narrowed it to two options and picked one without being sure which side of the distinction the stem put you on.
• Guesses. Anything where you eliminated and then chose without a reason.
• Questions where a later item might help. Regulatory papers repeat concepts across topics, and a Part XV question sometimes clarifies something you half-remembered on a Part XIV one.

That third category is real and underrated. Question 45 can remind you of the rule you needed at question 12.

What should you not flag?

Anything you answered confidently. Anything you found merely long. Anything you flagged out of general unease rather than a specific doubt.

Over-flagging is the more common failure. A candidate who flags twenty questions has not created a review list, they have created a second exam they do not have time to sit, and they will spend the final minutes skimming rather than resolving. Aim to flag between five and ten. If you are flagging more than fifteen, the issue is coverage rather than technique.

Why you should not revisit unflagged answers

Because the only possible outcomes are no change or a change for the worse. You answered those questions confidently at a point when you were fresher and less pressured than you are now, and going back over them at minute eighty-five with adrenaline running is not a better decision-making environment.

Change an answer only for a stated reason. Two reasons qualify: you misread the stem, most often a negative you skimmed past, or a later question reminded you of the rule. "It feels wrong now" is not a reason. Neither is a pattern in the letters.

Here is the opinion, and it is firmer than the usual advice. Discipline about what you revisit is worth more than any elimination trick. Candidates lose more marks to second-guessing correct answers in the last ten minutes than they gain from all the technique in this cluster combined, and the fix costs nothing: only open a question again if it carries a flag.

Flagging as a diagnostic in practice

Use the flag when you practice, not just on the day, and use it as data. At the end of every timed set, count the flags and check how many of them you got right.

The second row is the one to watch for. A candidate who is confidently wrong will not flag anything and will not review anything, and the error survives all the way to the exam. It is the strongest argument for reviewing your correct answers occasionally as well.

Building the habit

Flagging is trivial to do and surprisingly hard to do consistently, because under pressure the instinct is either to flag everything or to forget the feature exists. Practice it in every timed set until the sequence is automatic: answer, flag if unsure, move.

The concession: none of this is worth much if you are behind on the clock. A candidate arriving at the final stretch with unanswered questions should ignore their flags entirely and spend the time making sure every item has something selected. Flags are a luxury of good pacing, and pacing comes first. That is covered in time management in the exam.