An ORE Part 1 mock-exam error log should tell you what to practise next, not reproduce every explanation in the bank. Record the decision that went wrong, the reason it went wrong and the action that could change a later attempt. Include uncertain correct answers when the result concealed a weak understanding.
For the sitting on 14 and 15 October 2026, listed in the GDC's published announcement and checked on 3 October 2026, the useful output is a manageable final-review plan. This is a suggested method, not a validated readiness score or a prediction of the examination's questions.
Preserve the conditions of the mock
Write down the resource, date, whether the questions were unfamiliar and whether the attempt was timed or interrupted. Also distinguish independent answering from a session in which explanations or notes were consulted. The latter can be good learning, but it answers a different question about performance.
A percentage without those conditions invites misleading comparisons. A repeated paper may feel easier because its wording is familiar. A different provider's paper may sample different material. Neither observation, by itself, establishes improvement or deterioration in your underlying knowledge.
Keep the conditions in a single line. The error log should remain easy to revisit, rather than becoming a project that competes with revision.
Use four fields that lead to action
A practical entry contains the task, the error mechanism, the correction and the retest. For example: "Interpret the finding; confused the finding with its cause; explain the mechanism from the reference; attempt a different clinical presentation later."
The task is more precise than a broad specialty name. The mechanism explains what happened, rather than merely recording that the answer was wrong. The correction identifies an activity you can complete. The retest establishes how you will find out whether that activity helped.
Do not copy the whole question into an unauthorised external system. Use a brief learning description that respects the provider's terms and contains no patient-identifiable material. The log needs the principle, not an archive of proprietary questions.
Separate four common sources of error
A knowledge gap means the necessary concept was not available to you. An interpretation error means relevant knowledge was present but you answered a different question, missed a qualification or used the wrong part of the stem. A UK-context gap means the relevant professional framework needs checking. A pacing problem concerns how the decision process used the available time.
These categories can overlap. A learner may read slowly because a mechanism is uncertain, not because every stem requires a faster reading technique. Another may know the science but spend too long reconsidering an answer without finding new evidence.
Choose the main correctable mechanism rather than arguing over labels. If the next activity would be the same, the distinction does not need further administrative work.
Three fictional entries
A fictional candidate, Nisha, records a science error as "oral pathology weak". That label does not identify a next step. She replaces it with "could not connect the underlying process to the finding described" and schedules a short reference check followed by self-explanation.
Her second entry concerns a professional scenario. She selected a response familiar from a previous workplace without establishing the UK professional expectation. The correction is to inspect the relevant GDC standard, record its qualification and apply it to a different fictional situation.
Her third entry concerns an answer that was correct. Nisha changed it repeatedly and used more time than the task justified. She records the point at which further consideration stopped adding information, then tests a more disciplined review process in a later timed set.
These examples are constructed to demonstrate the method. They are not reports of candidate outcomes, official ORE items or evidence that a particular category occurs at a known frequency.
Prioritise recurring and consequential misunderstandings
Not every wrong answer deserves an extended entry. Prioritise a concept that repeatedly causes errors, an explanation you cannot defend or a neglected area identified against the official scope. Keep less important one-off mistakes available within the bank's own review system where possible.
Use the GDC Part 1 description, checked on 3 October 2026, to keep both applied science and the clinical and professional domains visible. A log built only from your favourite question category cannot identify material you have never sampled.
Avoid assigning numerical risk scores that suggest more precision than the observations support. A simple distinction between "needs action before the sitting" and "resolved enough for a later mixed check" may be sufficient.
Make the correction smaller than a chapter
For a mechanism error, choose a specific explanation and articulate it without looking. For a close differential, compare the information that would distinguish the options. For a professional rule, check the current source. For pacing, practise the exact behaviour that caused delay.
A correction should end with something you can demonstrate. "Read more" has no obvious stopping point. "Explain why the alternative does not answer this task" does. That boundary helps prevent a final-fortnight review from expanding into an entire new course.
When an explanation remains unclear after focused work, seek appropriate teaching rather than repeatedly copying the same answer. The point of the log is to reveal what is unresolved, not make unresolved entries look complete.
Retest with changed wording and context
An immediate repeat can check whether you remember an explanation. A later unfamiliar question is more informative about whether you can use the distinction independently. Where possible, vary the presentation rather than simply changing the order of the original answer options.
Keep the retest result brief. Record whether the original misconception recurred and what, if anything, still needs attention. If the answer is correct for the wrong reason, the entry is not resolved merely because the percentage improved.
Do not keep adding new entries without revisiting earlier priorities. A shorter log with completed learning actions can be more useful than an exhaustive collection of unresolved mistakes.
Use iatroX for the explanation stage
Per iatroX product information, September 2026, its question-linked Socratic Tutor can explore why a learner selected an answer and ask follow-up questions about the misconception. Adaptive sequencing and spaced repetition are designed to support subsequent practice. These are educational features, not a validated prediction of ORE success.
A useful session begins with Nisha explaining her original reasoning. The discussion identifies the missing distinction, and the log records only that distinction and the next independent task. The Tutor should reduce the ambiguity of the correction, not generate a longer record that she never reviews.
Keep any clinically important explanation subject to source checking. An AI-generated summary is not evidence that the underlying professional or scientific claim is correct.
Close the log before it becomes the workload
In the final days, review the small set of persistent issues and confirm examination arrangements. Do not try to reconstruct a complete history of every question answered. Your final record should explain the decisions you now make differently and the remaining uncertainties that deserve attention.
The most useful measure is not the number of entries written. It is whether the next unfamiliar attempt shows a better-supported decision.
Frequently asked questions
Should I record every incorrect ORE answer?
No. Prioritise recurring misconceptions, important unresolved explanations and gaps in coverage, while using the bank's own history for routine review.
Should correct answers appear in the error log?
Yes, when they were guesses, took disproportionate time or depended on reasoning you cannot explain. Correctness alone can conceal a learning need.
Can my error log tell me whether I will pass?
It can identify preparation priorities but cannot provide a validated pass prediction. Interpret mock results alongside their conditions, coverage and the nature of the remaining gaps.