1.2 Question Bookmarking and the Review & Edit Feature Mechanics

1.2 Question Bookmarking and the Review & Edit Feature

Official mechanics

The current GMAT exam is computer-adaptive and also permits limited review within each section. Keep the official interface rules separate from any preparation strategy:

1. Answer before moving on. You must submit an initial answer to the current question before proceeding. You cannot skip an unanswered item and return later.
2. Bookmark while working. You may mark questions you want to locate quickly at the end of the section.
3. Finish the first pass. After you have answered every question in the section, the Question Review & Edit screen appears if time remains.
4. Review any question. The screen lists every question and identifies your bookmarks. You may revisit bookmarked or unbookmarked questions.
5. Edit up to three answers. You may change no more than three responses in that section. Reviewing a question without changing the selected response does not use an edit.
6. Stay within the same 45-minute clock. Review time is not added to the section. If no time remains after the final first-pass answer, the review screen is unavailable.

These rules apply separately in Quantitative Reasoning, Verbal Reasoning, and Data Insights. The three-edit allowance resets for each section.

What an edit does—and does not do

Changing an answer cannot replace or reorder questions that were already delivered. The adaptive sequence has already occurred. The finalized response is nevertheless used in scoring.

Do not build strategy around invented claims about hidden scoring weights. GMAC specifically identifies several scoring statements as myths: an early wrong answer is not automatically weighted more heavily merely because it came first, and changing a response from correct to incorrect is not given an extra penalty beyond being an incorrect finalized response. Your practical objective is simpler: use an edit only when you have a concrete reason to believe the current answer is wrong.

Examples of concrete reasons include:

• discovering an arithmetic or sign error;
• noticing that you answered the converse of what the question asked;
• finding a sentence in the passage that directly contradicts your selection;
• recognizing that you treated a Data Sufficiency “definite no” as insufficient;
• realizing that you selected values in the wrong Two-Part Analysis columns.

General anxiety is not evidence. If a second reading produces no identifiable flaw, retaining the original choice is usually more disciplined than switching at random.

A practice-derived bookmark system

The following is a preparation method, not an official GMAC rule. Adapt it using your own practice data.

Priority 1: fast, objective checks

Bookmark a question when the setup is sound but one step is easy to verify quickly:

• a multiplication, fraction, or sign;
• a copied value from a table;
• a boundary word such as “at least” or “greater than”;
• a selected row or column in a two-part response.

These are strong review candidates because a short check can produce high confidence.

Priority 2: a genuine two-choice decision

In Verbal or Data Sufficiency, you may reduce the problem to two plausible choices without resolving the decisive distinction. Bookmark it only if you can state what you need to recheck—for example, the author’s scope, a quantifier, or a missing counterexample.

Priority 3: an unsolved time sink

If you made a reasoned guess because a problem would require a complete restart, a short review window may not be enough to solve it. Bookmarking every such item can crowd out faster, more reliable corrections.

Building a review buffer

No official rule promises that you will have time to review. A review buffer is created only by your pacing. During practice, record:

• time remaining after the final first-pass answer;
• number of bookmarks;
• number of reviewed items;
• number of answers changed;
• whether each change improved or harmed accuracy.

Those measurements tell you whether review is actually helping. For example, if you routinely rush the last four questions to save four minutes and then correct only one item, the trade may be negative. First-pass completion and accuracy remain the priority.

A starting experiment is to aim for a small buffer—perhaps two or three minutes—without forcing it. Adjust only after several full sections. Faster candidates may benefit from more review; candidates near the time limit may be better served by finishing the first pass carefully.

Worked scenario

David completes Quantitative Reasoning with 2:40 remaining and has bookmarked four questions:

1. Question 5: He sees that he used a negative root for a physical length. He changes the response. One edit is used.
2. Question 9: He recomputes an exponent and confirms the existing choice. No edit is used.
3. Question 14: He notices that he double-counted an overlap in a set problem. He changes the response. A second edit is used.
4. Question 18: He still lacks a workable setup for a difficult counting problem. With little time left, he preserves his reasoned first-pass guess.

David reviewed three questions but changed only two. The edit limit applies to changed responses, not merely to opening a question.

Common traps

1. The deferral trap: rushing the first pass because you assume review time will be available.
2. Over-bookmarking: marking so many questions that the list provides no useful priority.
3. Changing without evidence: switching because the original answer “feels too easy.”
4. Forgetting the clock: spending the entire buffer on one question while easier checks remain.
5. Treating three edits as a quota: you may change up to three answers; you are not required to use all three.

Confirm the current interface rules on GMAC’s official exam structure page before test day.