Choose a method that fits your goals, time, and fidelity. Then recruit participants who genuinely reflect your users, including those using assistive technologies.
Sessions: 30–45 minutes each; two researchers (moderator + note‑taker).
Recording: with explicit consent only; store securely; anonymize in notes/clips.
Environment: stable network for remote; quiet room and camera positioning for in‑person.
Debrief: 10 minutes after each session to log issues, severity, and hypotheses while details are fresh.
Scheduling & incentives: allow buffer time between sessions; compensate fairly and consistently per locale norms.
Bias Guards
Use consistent introductions and prompts.
Randomize task order when possible.
Avoid leading language and facial reactions.
Separate observation from interpretation in notes.
Data Hygiene
Label notes by task and timestamp.
Capture exact wording for quotes; tag with theme codes.
Mark assistance moments; they matter for success definitions.
Track environment (device, browser, AT) for reproducibility.
Summary: Method drives what evidence you can collect; participants determine how credible it is. Define procedure and bias guards up front, recruit inclusively, and plan accommodations so all users can participate effectively.