Applied Evidence
Guide five of seven

What a qualitative study can tell you

It does not measure whether a treatment works. It finds out what something is like, in enough detail that you recognise it. That is a different question, with answers a trial cannot produce.

Written and reviewed by Reza D, OTR/L, MOT, CPACC · Published 10 September 2026 · Free to read

A qualitative study does not measure whether a treatment works. It finds out what something is like, in enough detail that you recognise it. That is a different question, and it has answers a trial cannot produce.

What question does a qualitative study answer?

It answers what, how and why, for a specific group of people in a specific situation. What did they find difficult. How did they explain it to themselves. Why did they stop coming. A trial can tell you that half the participants dropped out. Only a qualitative study can tell you what they said when asked.

The output is themes rather than numbers. A theme is a pattern the researchers found across their interviews, stated in a way you can check against the quotes they publish. That is the evidence: not a p value, but whether the interpretation is supported by what people actually said.

Comparison
What the common qualitative approaches are for.
ApproachWhat it producesWhen it is the right choice
Grounded theoryA theory built up from the data, describing a process people go through.When you want to understand how something unfolds over time and no adequate explanation exists yet.
PhenomenologyA description of what an experience is actually like from the inside.When the point is the lived experience itself, not the process or the service around it.
Thematic analysisPatterns across a set of interviews, organised into named themes.The general-purpose choice. Most of the qualitative papers here use it.
Photovoice and participatory methodsAccounts produced with participants rather than about them, often using their own images.When the group is usually spoken for, and the point is to hear them directly.

Why does every qualitative paper here rate very low?

Because the certainty rating on this site answers one question: how much weight can this single paper carry as evidence that a treatment works. A qualitative study is not built to answer that, so it rates at the bottom of a scale it was never on.

That is a limitation of the rating, not a criticism of the research. The certainty guide explains where the rating comes from and why nothing here is ever rated high. Read very low on a qualitative paper as a reminder of what the paper is not claiming, rather than as a mark against it.

What can you actually take from one?

Three things, reliably.

  • Language. The words patients use for their own problem are the words that work in the clinic. Qualitative papers publish them verbatim.
  • Barriers you would not have guessed. Usually the reason an intervention fails is not the intervention.
  • What to ask. A theme is a question you can put to your own patient to see whether it applies to them.
Worked example
Barriers · physical therapyOffice workers' perspectives on implementing active breaks supported by a digital intervention to alleviate spinal painBrazilian Journal of Physical Therapy, 2026 · qualitative studySeventeen office workers with persistent spinal pain agreed that short breaks would help. They also said workloads and interruptions make them almost impossible to take. No trial of break frequency would have surfaced that, and it is the thing that decides whether the advice survives contact with a real job.Read the summary
Worked example
Language and mechanism · occupational therapyAn embodied journey towards recovery: a grounded theory of mental health service users' experiences of a nature-based interventionScandinavian Journal of Occupational Therapy, 2026 · grounded theoryTwenty people who had found talking therapy and medication unhelpful described a farm programme as working through the body and the senses instead. That contrast is the finding. It tells you which patients might respond and what they are likely to say about it, neither of which appears in an outcome score.Read the summary
Worked example
Hearing directly from people usually spoken for'Let's just hope for the best': a photovoice study exploring the impact of Covid-19 restrictions on children with intellectual disabilitiesJournal of Applied Research in Intellectual Disabilities, 2026 · photovoiceThirteen children with intellectual disabilities described lockdown themselves, rather than being described by a parent or a clinician. They talked about lost routines and fear, and about finding comfort in sensory activities and special interests. A proxy-reported questionnaire would have produced a score and none of that.Read the summary

How do you tell a good qualitative study from a weak one?

Five questions, and none of them require statistics.

  • Do they show you the data? Themes should come with quotes. If you cannot see what people said, you cannot check the interpretation.
  • Do they say who is missing? Fourteen caregivers from one community is a finding about that community. A good paper says so.
  • Do they say when they stopped? Recruiting until no new themes appear is a defensible stopping rule. Recruiting until the funding ran out is not, but it happens, and honest papers admit it.
  • Is the researcher in the paper? A clinician interviewing their own patients gets different answers than a stranger would. Good qualitative work states the relationship rather than pretending to neutrality.
  • Do the conclusions stay inside the data? This is the common failure. A themes paper that ends by recommending the intervention has stepped outside what it can support.

What can it not tell you?

Whether the treatment works, how often, or for whom in general. Participants who describe a programme as helpful are describing their experience, not demonstrating an effect. They were not randomised, there was no comparison, and the people who found it useless are frequently the people who left before anyone interviewed them.

The other thing it cannot do is transfer. A study of nine Japanese therapists setting goals over video is about those nine, in that system. Whether it applies to you is your judgement, and the paper should give you enough detail to make it. If it does not describe the setting clearly enough for you to decide, that is a flaw.

If you are working out how much weight any single paper can carry, the guide on studies that cannot prove anything covers the other designs that answer a real question without answering that one. You can also browse every paper here for occupational therapy.


Other guides
01What the certainty rating on a research summary actually meansEvery paper on this site carries one of three words: moderate, low, or very low. Here is where each comes from, what it does not mean, and what it should change about how you read the summary underneath it.02MCID vs MDC: telling whether a change on an outcome measure is realTwo numbers decide whether a score has moved: one asks whether the change is bigger than the measurement error, the other whether it is big enough for the patient to care. They are not the same, and one is useless without the other.03Pilot, feasibility, protocol, case report: what a study that cannot prove anything still tells youA pilot, a feasibility study, a case report and a protocol each answer a real question. None of those questions is whether the treatment works — and reading them as if it were is the commonest way to misread the literature.04Scoping review, systematic review, meta-analysis: which one answers a clinical questionOnly one of them is designed to answer whether a treatment works. Telling them apart takes about ten seconds, and it changes how much weight the paper can carry.06Reading a validation study: the numbers that decide whether an assessment is usableA validation study asks whether an assessment measures what it claims, in the people you are using it on. It answers with a handful of numbers, and two of them decide whether the tool is usable at all.07"No significant difference" is not "no difference"It means the study did not find an effect. It does not mean there is no effect. Those are different claims, and only one of them is usually supported by the paper in front of you.

Written and reviewed by Reza D, OTR/L, MOT, CPACC. Guides are the only pages on Applied Evidence written by a person. Paper summaries are generated by software and marked as such. Last reviewed 10 September 2026.