1. Read for intent
The Word document supplied the syllabus objective, mathematical language and question evidence. Each item was analysed for the concept, representation and likely misconception it could reveal.
A teacher-facing guide to Argand Modulus and Argument Observatory: a six-stage, misconception-first interactive created from Liang Soon's Word document and packaged for Singapore Student Learning Space.
Modulus and argument are learned as measurable geometric quantities attached to a point, not as disconnected formulas. Dragging, radius and angle cues coordinate Cartesian and polar descriptions while preserving exact values.
The AI did not merely generate answers. It translated a static assessment document into an instrumented learning progression: learners predict, attempt, receive misconception-specific feedback, open a visual tutorial, retry, and leave semantic evidence that a teacher can inspect.
The Word document supplied the syllabus objective, mathematical language and question evidence. Each item was analysed for the concept, representation and likely misconception it could reveal.
Create an accessible Argand observatory with tap, drag and keyboard placement. Make modulus a radius, argument an oriented principal angle, require quadrant reasoning before inverse tangent, compare distances using squared moduli, and reconstruct exact Cartesian form from polar data.
The payload is self-contained and responsive, with touch targets, keyboard access, read-aloud support, closed overlays on launch and no dependency on a network library.
The supplied reference ZIP was treated as a contract. lib/xapiwrapper.min.js, lib/xAPI.js and the injected xAPI head block were copied byte-for-byte; only the learning payload was redesigned.
Ask students to estimate radius and angle visually before calculating, then explain any difference between the estimate and exact value.
Every activity was checked at desktop, mobile and narrow-phone sizes; incorrect-feedback, visual-tutorial and correct-retry paths were exercised; and the package validator confirmed six stages, root-level ZIP entries, offline assets and byte-identical xAPI integration.
MathematicsComplex NumbersAI GeneratedSLSArgand Diagram