How oral exam rehearsal with dictated transcripts builds composure

Oral examiners grade deliverable structure and composure, not only knowledge, and the distinction is why excellent students fail vivas. The examiner hears a response's shape: an opening that answers the question directly, a body that moves through the reasons in a recognizable order, and a close that steps back to the bigger claim. A candidate who knows the material deeply but delivers it as a wandering narrative leaves the mark scheme unfulfilled, while a less knowledgeable candidate with a clean arc can meet the rubric's structure requirements. The rubric rewards arrivals at the right points, not the path between them, and the path is what most candidates rehearse, which is why oral exam rehearsal must be practiced as speech rather than as thought.

The second factor is composure, which is partly recoverable through familiar speech. A candidate who has said the answer out loud ten times does not spend working memory on the retrieval, and the freed attention handles the examiner's interruption, the follow-up, the hard sentence. The candidate who has only thought about the answer must run the retrieval live, under pressure, with a listener watching, and the mental load accumulates until the pacing collapses. This is why silent rehearsal is the worst preparation for a viva: it trains the content's encoding without training the production channel the exam actually uses, and the unbuilt channel is what costs the second mark on an otherwise correct answer.

The fix starts with a simple reframing: an oral exam is a performance with a script you write yourself. The script is not memorized word for word, it is a set of structures you can rely on, the opening sentence, the signposts, the summary, and rehearsing with the script written out gives you a target. Dictation turns the rehearsal into a written artifact that you can edit, which is an advantage the mirror never offered. The transcript shows you the shape of what you said, so you can fix the shape before the examiner ever hears it, and the edited shape is the version the next rehearsal should start from, not the original one.

The core rehearsal is simple and uncomfortable: take a likely question, set a five-minute timer, and dictate a complete response cold, with no notes and no warm-up. The cold start matters because the exam starts cold, with the examiner's first question landing before you have found your voice. Dictating your way through the response captures the answer's first draft exactly as it would arrive in the room, including the false starts, the long pauses, and the sentences that trail into silence. Those defects are the material you will edit, and you cannot edit what you cannot see, so the cold run is not a performance test, it is the capture that makes the rehearsal room possible.

The five-minute limit is chosen because viva answers follow the same rhythm: about ninety seconds for the opening and the main claim, three minutes for the supporting structure, and thirty seconds for the close. Run the dictation without watching the transcript while you speak, because watching the words as they appear changes your speech into writing: your sentences become longer, your pauses become shorter, and you perform differently than you would in the room. Review the transcript afterward, not during. The first cold run is normally a mess, and the mess is the point, since the mess names exactly the places the next run needs to be built for.

Do one cold run per question at the start of the preparation, and the messiness itself is diagnostic in a useful way. The places where the answer is thin are the places the examiner will press, and the places where the pacing dies are the places your confidence is weakest. Mark the transcript's weak sentences and long silences, then dictate the response a second time within the hour, informing your brain that these are the areas where the answer needs support. The pattern of the two runs is the rehearsal diary that shows the improvement, and the diary is what the preparation's final week should read, not the transcript of the first attempt.

The edit pass works on three elements: openings, signposts, and pacing. The opening is the three-second window where the examiner forms the impression of the whole answer, so the opening should repeat the question in a way that demonstrates understanding: 'the question is whether the mechanism is consistent, and the answer requires separating the two pathways.' Fix each opening to a clean version, and then use the edited opening in every future rehearsal of that question, since the opening is the element that varies least between runs. A good opening buys the time you need for the rest, and a bad opening spends the examiner's attention before the content ever arrives, so the opening earns the first minute of every rehearsal.

Signposts are the auditory structure of the answer: 'first,' 'the second reason concerns,' 'this is important because,' 'the key limitation is.' They are what a listener uses to track your argument, and an examiner who can track the argument awards the structure marks. In the edit pass, insert the signposts where the transcript shows the answer changing direction, and remove the hedges that appear at the start of each new segment, the 'well' and 'I mean' that pad the transition. The transcripts show the hedges plainly, and they show the places where a signpost was missed, because the text box changes topic without warning.

Pacing is the third element, and the transcript shows it as run-on sentences, which signal speaking too fast, and as short fragments, which signal hesitation. Edit for the marks of the ideal answer: one claim per sentence, roughly ten words per sentence at the most, with the occasional short sentence for emphasis. This kind of edit is not about writing, it is about traffic control: it puts the pause where the listening needs it. After the edit, re-dictate the whole answer once, and the improvement in the sound of the rehearsal is often dramatic, even between just two runs, because the second run is speaking the edited structure rather than finding it live.

Build the question bank by working backward from the course's likely questions. Use the syllabus's topic list, the previous years' viva lists if the department shares them, and the questions your lecturers ask in class, and write each as a standalone prompt. For each prompt, dictate one five-minute answer cold, edit it, and store it as the bank's entry: prompt, edited answer, and the three strongest sentences that the answer contains. The bank's size is the preparation plan: twenty questions at five minutes each is about two hours of dictation plus the edit passes, which is a realistic scope for a graduate viva.

The bank turns into a rehearsal deck by the final week, when you can run through the prompts rapidly, one after another, at half the answer time. The rapid run is a different skill: it builds the ability to start cleanly, which is the skill the exam's second and third questions demand, since they come without a rehearsal gap. Also keep the bank's answers in your own spoken language rather than the edited written version, because the exam's answer must be sayable. A sentence that works on paper but cannot be said in one breath will fail in the room, and the spoken check is the filter that catches the mismatch.

The bank also guards against over-rehearsal, which is the specific risk of the prepared candidate. An answer rehearsed into every nuance can sound mechanical, and the examiner's follow-up, the 'what if the constraint changes?' can unnerve a candidate whose rehearsed structure was not flexible. To counter this, keep the bank's answers at the level of the argument's skeleton, not the wording, and practice one variation per question: run the same answer with a changed constraint, a limitation, an alternative view. The bank of skeletal answers gives the flexibility, the rehearsed openings give the composure, and the two together are the honest preparation.

The spacing plan for the final two rehearsals is deliberately simple: one rehearsal at ten days before the exam, one at two days before, and a mirror run the day before. The ten-day rehearsal is the diagnostic one, where you dictate the whole bank cold and note the weakest answers. The two-day rehearsal is the verification one, where you run only the weak answers plus one full pass over the strongest. The gap is what makes the memory durable: the first rehearsal's weak spots, left alone for a week, get rebuilt by sleep and consolidation, and the second rehearsal finds them weaker to the frustration of the unprepared student, but the method knew they would be weak and planned for it.

The mirror run is the last preparation and it differs from the others: no notes, no recorder check, a standing position, a mirror, and the answers delivered aloud with the printed prompt read like the examiner's card. The mirror run serves two purposes: it builds the posture and eye contact habits that the examiner reads as composure, and it removes the crutch of the transcript for the final moments. A candidate who has rehearsed ten times with the recorder becomes dependent on the recorder, and the mirror run breaks the dependency a day before the exam, when the dependency is the only thing left to fix. Keep the mirror run to the weak answers and one strong answer, no more than thirty minutes, because fatigue the day before is counterproductive.

After the mirror run, do one final act: dictate a short list of the five things you must remember in the room, the opening line, the signposts, the one number or example that anchors the strongest answer, the pause before answering, and the closing sentence. Then stop rehearsing. The list is the safety net that replaces rehearsal, and the student who can say 'I am done, the list is in my pocket' carries a calm that no amount of extra rehearsal would have provided, and the calm itself buys the composure the examiners are listening for on the day, when the nerves arrive anyway.