A teleprompter for student video presentations — the limit, the rubric, the nerves

The assessed video presentation has quietly become a standard assignment: a five-minute pitch for a business module, a recorded talk for a graduate seminar, a spoken summary of a dissertation. The brief usually has two hard edges — a time limit with marks attached, and a rubric line about delivery or engagement — and neither is anything essay-writing skills prepare you for. An essay has drafts; a video has takes, and the takes have a deadline.

Nerves do specific, predictable damage. Speaking to a lens alone in your room, most people accelerate: a talk rehearsed at 140 words a minute comes out at 175, the ums multiply, and your eyes drop to your notes at every transition. None of it can be felt from the inside — which is why the fourteenth take feels different from the first and scores the same. The camera also flattens what a live room would forgive: in person, a glance at notes reads as thought; on video it reads as reading.

The time limit is discovered late. You rehearse in fragments, feel roughly on schedule, record a full run — and it lands at six minutes forty against a five-minute cap the rubric says costs a grade band. Cutting it down in an editor changes your pacing, which is the other thing the rubric marks.

And it is not only undergraduates. Doctoral students record three-minute thesis talks, job candidates record teaching demonstrations, and students presenting in a second language carry both loads at once: the content, and a delivery being marked in a language they think harder in.

The standard setups are all visible in the final file. Notes below the laptop camera read as downcast eyes for the whole take. A memorised talk means a blank in the middle, a restart from the top, and fourteen files named final in a folder the night before the deadline.

Timing is a phone stopwatch across rehearsals that never quite match the real run. Feedback is worse: you rewatch your own takes, cringing and none the wiser, or ask a flatmate, who says it was fine. Neither of you can name what changed between take three and take nine — and neither can the marker, who only ever sees the one file you chose.

The delivery advice on offer — slow down, pause more, look at the camera — is true and useless, because nothing tells you whether you actually did any of it on the take you are about to submit.

Some universities provide a media suite with proper lights and a booking sheet. What they do not provide is a way to rehearse before the booked hour — so the equipment is excellent and the delivery is take two.

The script sits on the same phone that is recording, under the lens, so reading and looking at your audience stop being different directions. Voice follow scrolls it as you speak: go blank, breathe, find the line — it waits rather than scrolling on into the next argument. Skip a sentence under pressure and it catches up. The restart-from-the-top spiral, where every stumble sends you back to zero, simply stops. And because you rehearse with the same script the take will use, the pacing you practised is the pacing you deliver.

For a strict limit, timed scrolling is the discipline: set the pace so the whole script fits the five minutes and the prompter holds you to it, with a progress bar showing the time remaining while you speak. If the script will not fit at a pace you can actually speak, you find out in the text — where cutting a paragraph is cheap — not at take nine.

Rehearsal gets numbers instead of feelings. Delivery metrics report your pace in words per minute, a filler-word count, and how much of the script you covered, measured on the device from the take itself. Eye contact analysis reports how much of the take you spent looking at the lens, your longest look away, and how steady the framing was — and when it cannot find a face it reports zero rather than guessing.

The numbers are the antidote to the advice: slow down becomes a words-per-minute figure you can aim at, look at the camera becomes a percentage that moved since yesterday. And the feedback is private in the plainest sense — computed on your phone, from your recording, shown to you. No platform account, no upload for analysis, nobody else watching the bad takes.

A marketing undergraduate has a five-minute assessed pitch: structure, pacing and audience engagement on the rubric. Her script, pasted from her essay plan, reads long, so she cuts it until timed scrolling fits it at a pace she can speak. First recorded rehearsal: 5:41, 21 fillers, eye contact 61 per cent — with the longest look away exactly where the revenue figure needed conviction.

Two more passes the same evening. She rehearses the weak paragraph with voice follow waiting through her pauses instead of punishing them, and the numbers move: 4:56, nine fillers, eye contact 84 per cent. The take she submits is the fourth, not the fourteenth, and the saved hours go to the written component that carries the other half of the mark.

A term later the next assessed video starts from routine rather than dread: script, timed pass, two rehearsals, submit. The nerves do not disappear. The fourteen takes do.