quality-checks-diagnosis — Claude Code skill for surveys
quality-checks-diagnosis helps choose the right survey check and interpret failures. Use it when you need to distinguish inattention from unclear instructions or a weak experimental treatment.
no license ★ 240 updated 2026-09-18
npx skills add franklee16/academic-research-skills --skill quality-checks-diagnosis -a claude-codeWhat quality-checks-diagnosis does
The skill separates attention checks, comprehension checks, and manipulation checks. It connects each failure type to a different part of the study to review: the respondent’s data, the instructions, or the treatment design. A failed comprehension check can point to confusing instructions rather than an inattentive respondent.
It recommends 1–2 attention checks, 1–2 comprehension checks before items about the study construct, and one manipulation check per treatment arm. It also advises counterbalancing check positions so their place in the survey does not skew failure rates. This skill focuses on choosing and interpreting checks, not on revising question wording or planning the operational flow of a pilot.
What it helps with
- choose a check type for a survey
- diagnose why respondents fail checks
- test whether respondents understand instructions
- assess whether an experimental treatment had its intended effect
- decide how many checks to include
Example prompts for the agent
- Help me choose checks for my survey: I’m unsure whether I need attention checks or comprehension checks.
- Respondents are failing a check. Identify what the failure might mean and what I should revise.
- How can I tell whether respondents noticed the difference between the experimental conditions?
How to install quality-checks-diagnosis
1Skills CLI (Vercel)
npx skills add franklee16/academic-research-skills --skill quality-checks-diagnosis -a claude-code2Manually: put the files in place
~/.claude/skills/quality-checks-diagnosis/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/quality-checks-diagnosis/
1Skills CLI (Vercel)
npx skills add franklee16/academic-research-skills --skill quality-checks-diagnosis -a codex2Manually: put the files in place
.agents/skills/quality-checks-diagnosis/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add franklee16/academic-research-skills --skill quality-checks-diagnosis -a cursor2Manually: put the files in place
.cursor/skills/quality-checks-diagnosis/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install franklee16/academic-research-skills running-surveys-experiments/quality-checks-diagnosis/SKILL.md2Manually: put the files in place
.github/skills/quality-checks-diagnosis/SKILL.md1Skills CLI (Vercel)
npx skills add franklee16/academic-research-skills --skill quality-checks-diagnosis -a gemini-cli2Platform CLI
gemini skills install https://github.com/franklee16/academic-research-skills --path running-surveys-experiments/quality-checks-diagnosisManually: put the files in place
.devin/skills/quality-checks-diagnosis/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about quality-checks-diagnosis
How do I install quality-checks-diagnosis in Claude Code?
Run npx skills add franklee16/academic-research-skills --skill quality-checks-diagnosis -a claude-code. The skill file is at ~/.claude/skills/quality-checks-diagnosis/SKILL.md.
What does quality-checks-diagnosis do?
It distinguishes checks for attention, instruction comprehension, and experimental manipulation, then helps connect failures to the part of the study that may need revision.
How can I tell why respondents are failing a check?
The skill ties the failure to the check type: an attention-check failure can raise concerns about response quality, while comprehension failures can point to unclear instructions. A failed manipulation check can mean the conditions did not create a noticeable difference.