Quest Ladder
TeachFORGE.ai
Quest Ladder / Hotspot and labelling
Hotspot on image and diagram labelling
When the answer is a place on an image rather than a word in a box.
Point at it. That is the question.
The two formats
| Format | The learner | Settings |
|---|---|---|
| Hotspot on image | Clicks the correct region | Number of hotspots 1–8, default 1 |
| Diagram labelling | Names the marked parts | Number of labels 2–12, default 4 |
They answer different questions about the same picture. Hotspot asks where is it; labelling asks what is that. Anatomy, maps, circuit diagrams and annotated graphs all want one or the other, and often both in sequence.
Hit radius
A hotspot is a point with a tolerance, and Hit radius (px) on the correct answer step is that tolerance. It is the single setting that decides whether the question is fair.
Too small and a learner who knows exactly where the structure is misses by three pixels. Too large and two adjacent regions overlap, so a wrong click scores. Set it against the real picture at the size a learner sees it, not against the full-resolution original.
Check it on a phone. A radius that is comfortable with a mouse can be unusable with a fingertip, and a learner on a small screen is not getting an easier question — they are getting an impossible one.
Labelling
Each label is a slot with its own marks, so a learner who names six parts of eight gets six parts of the credit when Marking mode is weighted. Set the label text carefully: it is compared with the answer policy's rules, so decide in advance whether a common abbreviation should pass.
Number of labels goes to 12, which is more than most diagrams should carry. A picture with twelve labels is usually two questions wearing a coat, and splitting it gives you better data about which half the class does not know.
Getting the picture in
Both formats need an image, and it does not have to be a photograph you uploaded. A diagram drawn with the platform's own tools is sharper at every screen size, translates without a second version, and can be edited later when it turns out to be slightly wrong.
Visuals without uploading images covers the ways to do that. Where you do need a real photograph, upload the largest clean version you have and check the hit radius against the displayed size.