“When I tell a project team that their appointment system is impossible for shift workers, I am not giving them a vibe. I am reporting a pattern from my life and my colleagues’ lives.”
— a community mediator, composite voice
Lived experience is evidence. It is evidence about what people encounter, how systems are navigated, what harms feel like, where formal rules do not match practice, and which categories fail to describe reality. It is not the same kind of evidence as a survey, an administrative dataset, or an audit study. That difference is not a ranking. It is the beginning of good research design.
The composite voices in this article are role-attributed, with details merged and changed. They are included to show how different kinds of evidence speak, not to stand in for all experience.
“My Story Shows the Mechanism”
“The form said the service was open to everyone. The barrier was not the sentence on the website. It was the callback at 10:30, the German-only letter, and the assumption that if you miss one appointment you are not serious.”
— an advice worker in a small city, composite voice
Testimony often reveals mechanisms: the steps through which exclusion happens. A survey might show that a group uses a service less often. Administrative data might confirm lower take-up. A testimony can show the chain: inaccessible appointment times, unclear letters, fear of losing benefits, no childcare, and a staff tone that makes people stop trying.
This is why personal experience counts as evidence. The German question “Zählt persönliche Erfahrung als Evidenz?” deserves a clear answer: yes, when treated as situated knowledge rather than as a universal estimate. It can show what happened, what it meant, what people noticed, and what questions researchers should ask next.
The limit is scale. One person’s account cannot tell us how common the mechanism is. It can tell us that the mechanism exists and that it deserves investigation.
“An Interview Lets Me Explain the Order Things Happened In”
“If you ask me whether I trust the office, I may say no. In an interview I can tell you why: first the lost document, then the public question at the counter, then the letter that contradicted the phone call.”
— a participant in a service evaluation, composite voice
Interviews are useful when sequence, meaning, and interpretation matter. They allow follow-up questions. They let participants explain contradictions. They can show why two people with the same formal outcome experienced the process differently.
In research terms, interviews are qualitative evidence: they prioritize depth, context, and meaning over numerical representativeness. They are not informal chats. Good interviews require consent, careful question design, confidentiality, skilled listening, and analysis that looks for patterns across accounts rather than only memorable quotes.
Interviews can reveal what surveys miss: informal rules, fear, shame, workarounds, hidden costs, emotional labour, and the difference between access on paper and access in practice. They can also reveal where researchers’ categories are wrong. A survey may ask whether someone “used a service.” An interview may show that the person attended once but left because interpretation failed, which is not meaningful access.
“Ethnography Catches the Room, Not Only the Answer”
“In meetings, people say the consultation was open. Then you watch who speaks, who translates quietly, who leaves early, and whose comment becomes a decision.”
— a participatory researcher, composite voice
Ethnography involves observing social life in context. In civil-society research, that might mean observing a waiting room, a public consultation, a workshop, a neighbourhood meeting, or a service pathway. It can show interaction: who is deferred to, where rules are improvised, what is not written down, and how power appears in ordinary routines.
Ethnography is not superior to a survey. It answers a different question. It can show how a process works while it is happening. It can identify the gap between formal design and lived practice. It can also help explain why a program that looks accessible in documents feels inaccessible in the room.
Its limits are real. Observation is time-intensive. The researcher’s presence can change behaviour. Findings are shaped by site selection and interpretation. That is why ethnographic claims should be precise: “In these observed settings, this pattern appeared,” not “this is how all institutions behave.”
“A Survey Lets Us See Distribution”
“When the same question is asked of many people, we can stop treating the loudest example as the whole picture.”
— a survey researcher, composite voice
Surveys are strong when the question is distribution: how common something is, which groups report it more often, whether it changes over time, and how variables relate to each other. A well-designed survey can show that an experience is not isolated. It can also show variation within a group, which is often politically important.
Surveys need careful sampling, accessible language, respectful categories, and transparent uncertainty. A survey of an organization’s newsletter list cannot represent a whole city. A survey only in German may miss people the organization claims to serve. A question that combines too many issues will produce unclear answers.
Surveys are also limited by what respondents can know and report. People may not know why a landlord did not reply, why a school placed a child in a track, or why an application failed. They can report what happened to them and how they interpreted it. Researchers must distinguish those levels.
When organizations combine stories and statistics, surveys can estimate pattern while testimony and interviews explain mechanism. The statistic says, “This is not rare.” The qualitative evidence says, “This is how it works.”
“Administrative Data Shows the System’s Trace”
“The database tells us who got an appointment. It does not tell us who gave up before the appointment existed.”
— a municipal analyst, composite voice
Administrative data is produced by systems: schools, housing offices, Jobcenters, health insurers, courts, advice centres, or service providers. It can be powerful because it records actual transactions rather than only memories or perceptions. It can show waiting times, application outcomes, repeat contacts, regional differences, and procedural bottlenecks.
But administrative data usually reflects the system’s categories. It records what the institution needed to process, not everything people needed to live. It may miss those excluded before entry. It may record nationality but not racialization, appointment attendance but not why someone missed it, case closure but not whether the problem was solved.
This is why administrative data should not be treated as the neutral master record. It is evidence of what a system sees and stores. That is valuable, but partial.
“Participatory Research Changes Who Gets to Interpret”
“We were tired of being asked for stories and then seeing conclusions we did not recognize.”
— a member of a community advisory group, composite voice
Participatory research involves affected people in shaping research questions, interpreting findings, designing tools, or deciding how results are used. It is not the same as adding one quote after the analysis is finished. Done seriously, it changes what counts as a relevant question.
This matters because expertise is distributed. Researchers may understand sampling, consent, and analysis. Community members may understand risk, language, access, and the meanings of categories. A survey option that looks neutral to a researcher may feel stigmatizing to respondents. An interview question may be technically clear but emotionally unsafe.
Participatory work still needs method. Community involvement does not automatically make findings representative. Researchers still have to explain who took part, how decisions were made, what conflicts appeared, and what limits remain.
Combining Evidence Without Flattening It
Organizations often ask whether they should use stories or statistics. The better question is: what claim are you trying to make?
If the claim is that a barrier exists, testimony may be enough to identify it. If the claim is that the barrier is widespread, a survey or administrative pattern may be needed. If the claim is that unequal treatment occurs at first contact, an audit or correspondence study may be appropriate. If the claim is that a program changed outcomes, evaluation design and comparison thinking matter. If the claim is that a public narrative harms people, media analysis and community testimony may both be relevant.
Community Voices Media is the stronger lane for longform testimony and the craft of public storytelling. Civic Futures Lab is the stronger lane for using stories ethically inside campaigns. The Institute’s contribution is to keep the evidence claim attached to the method that can support it.
No evidence type is universally best. Bad surveys can mislead. Extractive interviews can harm. Administrative records can erase. Testimony can be overgeneralized. Participatory processes can reproduce local power if they are rushed or symbolic.
Good evidence work does not ask lived experience to become a survey before it is believed. It asks what that experience shows, what else we need to know, who should help interpret it, and how to make claims that are strong because their limits are visible.
