Decision tree: Which Decode method fits my question
Choosing the right research method starts with understanding what you want to learn and what type of evidence you need.
Use the guidance below to decide whether you need qualitative or quantitative research, moderated or unmoderated research, live or AI moderation, and monadic or sequential designs.
1. Which Decode Method Fits My Question?
Start with your research question, rather than choosing a method first.
Step 1: What do you want to learn?
If you want to explore:
- Why people behave a certain way
- What they think or feel about an experience
- Their motivations, needs, or barriers
- Unexpected reactions or issues
- An area you don't fully understand yet
→ Consider qualitative research.
If you want to measure:
- How common a behaviour or opinion is
- Preferences between options
- Differences between groups
- Scores or ratings
- Changes over time
→ Consider quantitative research.
Step 2: Are you testing defined options?
If you already have concepts or options to evaluate:
| What you want to learn | Consider |
|---|---|
| Compare two options | A/B Test or Preference Test |
| Compare or rank several options | Preference Test or Ranking |
| Evaluate combinations of features or prices | Conjoint |
| Evaluate one concept independently | Monadic Design |
If you don't have defined options and need to explore the problem first, consider AI Interviews or Live Interviews.
Step 3: Are you studying behaviour or perceptions?
Behaviour
If you want to understand whether participants can find, notice, or complete something, consider relevant User Research methods such as:
- Task Testing
- First Click Testing
- Five Second Testing
- Prototype Testing
Perceptions
If you want to understand whether participants understand, trust, like, want, or prefer something, consider:
- Quantitative research
- Open-ended questions
- AI Interviews or Live Interviews
Step 4: Do you need depth or measurable results?
Need depth?
→ Choose qualitative research, then decide whether you need moderated or unmoderated research and whether moderation should be live or AI-based.
Need measurable results?
→ Choose quantitative research and plan the appropriate sample size and quotas.
Step 5: Do you need behavioural or emotional signals?
If participant responses alone aren't enough, you can add relevant measurement layers such as facial coding, voice tonality, or eye gaze, where available and appropriate.
These are additional measurement layers and do not replace the underlying research method.