How to review a missed question so the next attempt changes
By Ezra Menon, MD, MPH
Reading an explanation can make a question feel obvious. The more useful test is whether you can identify what happened during your attempt and change that process on a fresh question.
Here is a review routine you can use after a missed answer, a guess, or a correct answer you cannot explain.
First, reconstruct the attempt
Before reading the explanation, write down what you thought the question asked and why your choice seemed right. Include the moment you became uncertain or changed your answer.
If you already read the explanation, separate what you remember thinking from what you reconstructed afterward. You want a record of the original decision, not a cleaner version built around the key.
Find the detail that decides the question
Look for the qualifier: most likely, directly supported, next, initial, or EXCEPT. Then name the fact, passage detail, result, or relationship that separates the best answer from your choice.
For example, an experiment showing less phosphorylated protein does not by itself prove that a compound binds that protein directly. If your answer added direct binding, write down that extra inference. "I missed the experiment question" is too broad to help.
Diagnose the first failure
Content means you did not know or understand a needed concept. Recognition means you knew it but did not recognize when to apply it. Reasoning means you made an unsupported inference or compared the answers incorrectly. Execution means a missed word, unit, calculation, or time decision disrupted an otherwise workable approach.
These categories can overlap. Start with the earliest failure you can support with evidence. Two people can choose the same wrong answer for different reasons, so the answer letter alone does not diagnose the problem.
Choose a repair you can actually practice
Match the action to the cause.
If the concept is missing, review it and explain it without looking. If recognition failed, identify the cue and practice applying the concept in a different setting. If the inference failed, state the evidence you need before making that claim. If execution failed, build a narrow check into the relevant step.
Replace "be more careful" with something observable: "Before calculating, I will write the requested quantity and units." Replace "review experiments" with "For each mechanism claim, I will check whether the experiment measured that mechanism or only a downstream result."
Use the change on a fresh question
Redoing the same question can show that you remember the explanation. A new question testing the same skill gives you a better opportunity to see whether your process changed.
You do not need to find a matched question immediately. When you encounter a new question where the same problem could happen, try the change and note what you did differently. If the error returns, refine the repair. If the new question exposes a different gap, record that separately rather than forcing it into the old label.
Keep the error log short enough to use
Keep a question reference, the specific error, the evidence for your diagnosis, and one repair. Add what changed when you encounter a relevant new question. You do not need to copy an entire copyrighted question or explanation into the log.
Example: "Experimental interpretation. Inferred direct binding from a downstream change. Next time, match each claim to what was measured. Retest with a fresh mechanism-versus-observation question."
Try the worksheet
The free "Why you missed that question" worksheet puts this process on one page and includes an original MCAT-style worked example on the second page. No email is required.
Explore tutoring for MCAT, USMLE Step 1, USMLE Step 2 CK, or shelf exams.