Date: 29 September 2026. This is an author simulation and local file/browser review, not a classroom trial, user interview, independent accessibility audit, live site test, funding review or measured learning outcome.
Teacher path
I opened Day 1 and Card A as if preparing a 25-minute class. The script gives a 2-minute opener, 5-minute model, 6-minute pair analysis, 7-minute individual brief and 5-minute exit. A learner who asks for a “live app” is asked who updates occupancy and what the visitor really needs before travel. I could put need-map on a desk or read its full text. On Day 5, Card E models a criterion before the independent new festival check; the staff key stays separate until a first response is collected. A real teacher may need more time after observing the class.
Learner path
I completed Check A before viewing its key: the need is reliable programme-pickup information, a dated twice-daily page is within the team's capacity, a “live stock” claim has no count process, and a planned trial could test whether a reader finds the last update time and pickup point. I then completed Check B: select a topic → match approved tags → display descriptions; staff review tags weekly. The simple filter has evidence from existing tags, while the proposed ML recommender lacks training/evaluation/maintenance evidence. I explicitly labelled criterion thresholds as proposed, not results. This adult author desk simulation checks clarity and alignment, not student comprehension.
Supporter/funder path
The README leads to all ten clean cards, 20 optional contexts, two public checks, separate key, five print/text/tactile aids, the offline lab, source crosswalk and rights ledger. The pack names the current QCAA v1.4 after the v1.3 request and records the v1.4 cover/header date discrepancy. The crosswalk gives eight Topic 1 row positions and clear partial limits. A funder can inspect file hashes and rights, but no real participant, accessibility conformance, employment benefit, privacy compliance determination or classroom effect is claimed.
Asset and interaction inspection
I generated five original A4 SVG/PDF mats and rendered all five PDF pages at 700-pixel height into a contact sheet. Titles, four-step diagrams, worked examples and tactile directions remained on-page with no observed clipping; PDF text was selectable. Their full text repeats all meaning without reliance on colour or arrows.
I opened the offline framing lab in a local headless browser at desktop (980 × 1240) and narrow (390 × 900) widths and visually inspected both screenshots. Its statement controls remained readable at both widths. Via the browser's actual JavaScript runtime, I selected the four correct categories and clicked Check my sort (reported 4/4 with per-row reasoning); switched from art-fair map to cinema (four new controls appeared); clicked with no choices (reported 0/4); and confirmed paper self-check text. This exercises the interface, but no actual learner used it. It stores no response and loads no external script.
Critique and next work
- The ten cases are diverse but fictional. The full QCAA topic expects a school-selected technology context and deeper study of existing solutions, emerging technologies, user needs and impacts using appropriate evidence.
- The lab is teacher-authored; its use does not demonstrate a student's programming skill or a completed digital solution. Later Unit 1 work must address user experiences/interfaces, algorithms, programming and generation/evaluation of a prototype.
- Scope and impact criteria here are planned tests. Before a real service is built, its team needs genuine user research, consent/permissions, accessibility testing and operational update processes. This pack cannot substitute for them.
- The key is public. If learners have already seen it, a teacher can write another fictional case while retaining the same reasoning target; the checks have no secure-exam status.