Lesson 8 — Visual, Responsive, and Accessibility QA

Duration: 120 minutes

Included delivery track: Follow LAB_SETUP.md for the supplied files and local simulation. Required activities use these materials; connected deployments are optional extensions. Instructor debrief: answer key. Real OAuth, cloud releases and payments are simulated or discussed through evidence packets. Deck: 08-visual-responsive-accessible
Lab: Digital Showroom Visual Review

Learners will be able to

inspect mobile, tablet, and desktop interfaces for layout and interaction regressions;
combine visual comparison with semantic accessibility checks;
identify when a model’s visual impression requires a deterministic or human follow-up; and
write inclusive, user-impact-based UI findings.

Instructor preparation

Use the local launch preview, R7 and both candidates. Prepare matched-viewport captures, computed color inspection and keyboard checks. Theme-driven heading changes are intentional. There is no menu, modal, missing-label fixture or focus trap. An instructor-prepared axe scan is optional; the core targeted check is manual keyboard/semantic verification.

Agenda

Time	Facilitation
0–10	Arrival: display a subtle visual diff and ask what a pixel comparison cannot tell us.
10–25	Teach responsive checks, visual baselines, semantic HTML, keyboard flow, accessible names, and contrast.
25–40	Demo desktop-to-mobile inspection plus a keyboard/semantic check; use an optional preconfigured axe scan if available. Explain why a green scanner is not accessibility certification.
40–50	Guided practice: ask AI to describe a screenshot, then verify its claims with DOM, keyboard, and viewport evidence.
50–60	Break.
60–95	Teams complete a visual review at three viewports, with a keyboard-only pass and a targeted accessibility check.
95–110	Critique findings for specificity: who cannot do what, at which viewport, under which condition?
110–115	Discuss approval of intentional visual baseline changes.
115–120	Exit ticket: write one accessibility test that is not a color-contrast scan.

Assessment

Teams submit before/after evidence, viewport/device details, and at least one semantic or interaction verification for every visual claim.