ux-research
Plan, conduct, and analyze UX research using proven methodologies. Use when planning user research, writing interview scripts, creating test plans, recruiting participants, analyzing qualitative data, or synthesizing findings into actionable insights. Covers generative, evaluative, and continuous discovery research with specific frameworks, scoring rubrics, and templates.
cuellarfr · 已有静态检查记录
安装与使用
运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。
npx skills add cuellarfr/design-skills --skill ux-research技能说明
name: ux-research description: Plan, conduct, and analyze UX research using proven methodologies. Use when planning user research, writing interview scripts, creating test plans, recruiting participants, analyzing qualitative data, or synthesizing findings into actionable insights. Covers generative, evaluative, and continuous discovery research with specific frameworks, scoring rubrics, and templates.
UX Research
Plan, conduct, and analyze user research that produces actionable insights for product design. This skill provides frameworks, methods, and practical guidance for every stage of the research process — from defining the right question to presenting findings that drive decisions.
When to Use This Skill
Use this skill when:
- Planning a research study (choosing methods, defining scope, writing a plan)
- Writing interview scripts or discussion guides
- Creating usability test plans with tasks and success metrics
- Recruiting participants and writing screener surveys
- Analyzing qualitative data (affinity diagrams, thematic analysis)
- Synthesizing research into personas, journey maps, or opportunity maps
- Setting up continuous discovery practices (weekly interviews, OSTs)
- Presenting findings and recommendations to stakeholders
Core Principle
Research is a tool for reducing risk, not a deliverable. The goal is always to inform decisions — not to produce documents. Research is not asking people what they like, not a political tool, and not science. It's applied inquiry that helps you solve the right problem.
Research Types
Choose your research type based on what decisions are in play:
Generative / Exploratory
When: Before you know what you're building. You need to define the problem. Methods: Interviews, field observation, contextual inquiry, diary studies, literature review Output: Problem definition, opportunity areas, personas, mental models
Descriptive
When: You have a design problem and need to fully understand the context. Methods: Contextual inquiry, fly-on-the-wall observation, journey mapping, card sorting Output: Current-state maps, workflow diagrams, mental models
Evaluative
When: You have potential solutions and need to test them. Methods: Usability testing, heuristic evaluation, A/B testing, click testing, tree testing Output: Severity-ranked issues, task completion rates, recommendations
Causal
When: A solution is live and you need to understand actual usage patterns. Methods: Analytics review, A/B testing, funnel analysis, surveys Output: Conversion data, behavioral patterns, quantitative validation
The Research Process
1. Define the Problem
Start with a clear problem statement. Base it on a verb that indicates outcome:
- Good: "Identify the barriers preventing users from completing onboarding"
- Good: "Evaluate whether the new checkout flow reduces abandonment"
- Bad: "Understand the user experience" (too vague)
- Bad: "Explore the problem space" (no measurable outcome)
Write 3-5 research questions. More than 10 means your scope is too broad.
2. Select the Approach
| You need to... | Use... | Time needed |
|---|---|---|
| Discover unmet needs | Generative interviews | 2-3 weeks |
| Understand context and workflows | Contextual inquiry | 1-2 weeks |
| Test a design concept | Usability testing | 3-5 days |
| Validate with numbers | Survey + analytics | 1-2 weeks |
| Quick expert review | Heuristic evaluation | 1-2 days |
| Prioritize features | Card sorting, buy-a-feature | 1 week |
| Understand navigation mental models | Tree testing, card sorting | 1 week |
| Track ongoing patterns | Continuous discovery (weekly) | Ongoing |
See references/research-methods-catalog.md for detailed method descriptions.
3. Plan and Prepare
Every study needs a written plan, even if it's brief. Use templates/research-plan-template.md.
Essential plan elements:
- Problem statement and research questions
- Methodology and rationale
- Participant criteria and recruiting approach
- Number of participants (see benchmarks below)
- Timeline and logistics
- How data will be captured and stored
- Who will participate from the team
Participant benchmarks:
- Usability testing: 3-5 per round (catch ~85% of issues with 5 users)
- Generative interviews: 5-8 per persona segment
- Card sorting: 15-20 for statistical patterns
- Surveys: 100+ for quantitative significance
- Continuous discovery: 1 per week minimum
4. Recruit Participants
Recruiting is the hardest part. Embrace it — good recruiting puts the quality in your research.
A good participant:
- Shares the goals and concerns of target users
- Embodies key characteristics (age, role, tech familiarity)
- Can articulate their thoughts clearly
- Is familiar with relevant technology at the same level as target users
Write a screener survey (see templates/screener-template.md):
- Define specific behaviors you're looking for
- Determine required tool knowledge and access level
- Set topic knowledge requirements
- Don't make it obvious who you're looking for (people will game it)
Recruiting tips:
- The importance of recruiting perfectly representative users is overrated
- Testing early and often matters more than perfect representation
- Don't be embarrassed to ask friends and neighbors for initial rounds
- Offer reasonable incentives ($50-100/hour)
- Keep the invitation simple
- Don't discuss the product beforehand
5. Collect Data
Data management:
- Use consistent naming:
Study-ParticipantName-YYYY-MM-DD - Get files onto a shared drive immediately
- Use familiar tools — technical difficulties kill research sessions
- Keep a research kit ready to go at a moment's notice
Core collection methods:
Interviewing — The most effective way to see the world as users do. See references/interviewing-guide.md for the full technique guide.
Quick checklist:
- Prepare an inter