Dictation

Why this shape

Listen and transcribe: spelling, grammar and listening together, the oldest language test and still one of the fairest.

An example. Listen and type the sentence exactly.

What it is

Media questions carry a clip, a video or an image as the prompt, and ask for a typed or chosen answer. The information is in the sound or the picture, not in the text, so the student must listen or watch.

Why it is needed

Listening is a skill that text questions cannot test, and a clip that plays once cannot be pasted into a chatbot. For languages, music, medicine, law and any field where a professional hears or sees something and responds, the media question is the closest a quiz gets to the real task.

When to use it

Use listen-once for real listening: a dictation, a comprehension, a phrase to translate, a heart sound, a mechanical fault. Use watch-and-answer for procedures and demonstrations. Use identify-the-source for style, period and authorship.

How to write one well

Keep clips short and clean, and say in the question what to listen for before it plays. If a clip plays once, tell the student so, and set the question so a single hearing is fair. Add a transcript for accessibility where the setting allows.

For dictation in particular. Read at a natural pace and accept the punctuation variants.

Mistakes to avoid

A clip so long that the question tests memory, not listening. Background noise that makes the answer luck. A question answerable from the text alone. No captions where a student needs them.

How EdProcto marks it

A chosen answer is marked by rule at once, and so is a dictation, against the accepted list. The typed answer of listen and answer, a caption and an identified source are open answers: each waits for you to grade against your model answer and rubric, or for Procto AI when you ask it on that answer. The clip's address is stored with the question; the play-once rule is held on the student's screen by a player of one button and a line that fills, with no pause and no going back.