Quest Ladder
TeachFORGE.ai
Quest Ladder / Written answers
Short answer, long answer and proof
When the thinking has to be expressed rather than selected, and what that costs you in marking.
The most honest formats, and the most expensive.
The three formats
| Format | Length | Marked by |
|---|---|---|
| Short answer | A few words. Maximum words in answer runs 1 to 40, defaulting to 8. | The engine, against your acceptance rules |
| Long answer (essay) | Paragraphs. Key points to assess runs 3 to 8, defaulting to 4. | You, in the marking queue |
| Proof | A structured argument. Key points to check, same range and default. | You, in the marking queue |
The word limit on short answer is doing real work. Set to 8 it keeps the answer inside what automatic matching can fairly judge; raise it to 40 and you have an essay being marked by string comparison, which will be wrong often enough to matter.
Making short answer markable
Short answer is automatic, so everything depends on the acceptance rules being generous enough to pass a correct answer and tight enough to fail a wrong one.
- Other acceptable answers, one per line, for synonyms and alternative phrasings.
- Optional keywords, where any one of them may count — useful when several words would each show the understanding.
- Text match mode set to lenient, unless spelling is the point.
Run a short-answer type past a real class before you build twenty more of the same shape. The delivery report's Common wrong answers shows what learners actually typed, and the first pass almost always reveals two or three correct answers your list did not have.
What arrives in the marking queue
Long answer and proof go to Marking. For each one you see the marking rubric you wrote, the learner's answer, a Mark (out of) field and an optional Feedback box.
You also see the attempt context — hints used and their penalty, the learner's own confidence, time spent, which ladder step it came from — and the composed total updates as you type, so you can see what the mark will actually become once the deductions apply.
Key points are the rubric
Key points to assess and Key points to check are not decoration. They are the rubric slots that appear beside the answer when you mark, and they are what makes marking thirty essays consistent at the end rather than drifting.
Write them as things that are either present or not — states the assumption before using it rather than good understanding. A key point you cannot tick reliably will not survive the second class.
Before you build a ladder of these
A ladder of thirty automatically marked questions is a delivery you press once. A ladder of thirty essays is an evening's work per class. Both are legitimate; only one of them scales.
A common shape is a ladder that is mostly automatic with one or two written questions at the points where the reasoning matters. You get the coverage and the marking stays finite.