Quest Ladder
TeachFORGE.ai
Quest Ladder / Marking
Marking: written, proof and graph-sketch answers
Everything waiting for a human mark, with enough context to mark it fairly.
The machine marks what it can. This is the rest.
What arrives here
Marking holds the answers to written, proof and graph-sketch questions that need your review. Everything else is marked automatically and never appears.
Multi-step and code-execution questions can also arrive, depending on how you set their marking up.
What you see per answer
- The marking rubric — the key points you wrote on the question type.
- The student's answer.
- Mark (out of {{max}}) and an optional Feedback box.
- The attempt context: hints used and their penalty, the learner's own confidence, time spent, which ladder step it came from, and any active-learning multiplier.
- Their active-learning responses, where the question had blocks.
A composed total preview updates as you type, so you can see what the final mark will actually be once the hint deductions and multipliers apply.
Look at the composed total before moving on, particularly on a question with heavy hint penalties. A raw mark of 8 out of 10 can compose to something considerably lower, and a learner seeing that without explanation reads it as an error.
Marking thirty consistently
The rubric is the thing that keeps the thirtieth script marked like the first. It comes from Key points to assess or Key points to check on the question type, which is why writing those well pays for itself here.
- Mark one question across every learner rather than every question for one learner. It keeps the standard steady.
- Read the attempt context before deciding. An answer produced without hints is not the same performance as the same answer produced with three.
- Write feedback for the ones where it will change something. Feedback on everything takes four times as long and is read less.
After marking
Marks do not reach learners until you release them. Clear the queue for a delivery, then use Release results to students on that delivery — see Results.
If the queue is consistently longer than your evenings, the cause is usually the ladder rather than the marking. A ladder with one written question at the point the reasoning matters gives you most of the insight for a fraction of the work; see Short answer, long answer and proof.