What is member checking?
Also called: Respondent validation, Participant validation, Member check
Member checking is a qualitative research technique where you take your interpretation of what participants said back to them, so they can confirm, correct or add to it before you treat it as a finding.
Updated
Where it comes from
Yvonna Lincoln and Egon Guba described member checks in Naturalistic Inquiry (1985) as a central way to establish credibility in qualitative work: the people you studied are the ones best placed to say whether you understood them. In product and UX research the same idea shows up whenever a researcher asks, "Did I get that right?"
Three ways to do it in user research
- In the session. Summarise back before you move on: "So the export step is the part you dread each month, because the numbers never match. Is that right?" This is the cheapest version and catches misunderstandings while the participant is still there.
- After synthesis. Send a short summary of the themes, written in plain language, to a handful of participants and ask what's missing or wrong. Keep it to one page and two or three questions.
- In a follow-up session. For high-stakes work, such as personas or a journey map that will guide a roadmap, walk a few participants through the draft on a call and watch where they push back.
What to do with the answers
Agreement is useful but not proof. People often agree to be polite, especially with a researcher from a company they pay. Ask for corrections, not approval: "What would you change?" gets better answers than "Does this sound right?"
Disagreement is data. If a participant rejects a theme, check whether you misread them or whether they see their own behaviour differently from how they act. Both are worth reporting. Note who disagreed and why, rather than quietly dropping the theme.
Limits to keep in mind
- Participants can only judge their own part. They can't confirm a pattern across 15 interviews they weren't in.
- Views change. Someone checked three months later may describe a workflow they have since changed.
- It costs participant time. Budget an incentive for follow-up calls, and don't ask the same people every time.
Go deeper
- User interviews guide
Planning interviews, including summarising back during the session.
- How to analyze interview data
Turning transcripts into the themes you check with participants.
- Research incentives guide
Paying fairly for follow-up and validation sessions.
- Interviews in Great Question
Schedule and run the follow-up sessions a member check needs.
Questions
Is member checking the same as triangulation?
No. Triangulation checks a finding against other sources, methods or researchers. Member checking checks it against the participants themselves. They work well together: triangulate across data sources, then check the result with the people it describes.
How many participants should you member check with?
You don't need everyone. For a typical interview study, going back to three to five participants across your main segments is enough to catch misreadings. Pick people whose interviews shaped the main themes.
Should you change a finding if a participant disagrees?
Not automatically. Look at what else supports the finding, such as behaviour you observed or product data. If the evidence still holds, keep the finding and note the disagreement. If the participant shows you misread them, revise it.
Run the study, not just the method
Great Question handles recruiting from your own customers, scheduling, incentives, interviews, tests and analysis in one place.
No credit card required.
Related terms
Full glossary- Triangulation
Triangulation is checking a research finding against more than one source, method or researcher, so you only act on it when different kinds of evidence point the same way.
- Social desirability bias
Social desirability bias is the tendency for research participants to answer in ways that make them look good, overreporting behaviour they think is approved of and underreporting behaviour they think is frowned on.
- Inter-rater reliability
Inter-rater reliability is the degree to which two or more researchers, coding or rating the same data independently, reach the same answers. It tells you whether a codebook or rating scale produces consistent results, or depends on who happens to apply it.