Home/Insights/Methodology reference/Descriptive research design
Descriptive research design
It characterises what is there. Without a comparison group it cannot support any statement about why, and the discussion is where that gets forgotten.
What it covers
Case reports, case series, surveys of prevalence, registry descriptions and audits. The output is a characterisation: who, what, how many, where, when.
Descriptive work is genuinely valuable. New conditions are usually recognised first in case reports, and service planning depends on prevalence data rather than on effect estimates.
The hard limit
There is no comparison group, so there is no counterfactual. A case series in which 80% of patients improved after a treatment tells you nothing about the treatment, because the proportion who would have improved anyway is unknown.
This is the single most common over-claim in clinical writing, and it survives peer review more often than it should.
Selection determines what the description means
A case series from a tertiary referral centre describes referred patients, not the disease. Consecutive enrolment is substantially stronger than a selected series, and whether enrolment was consecutive should be stated explicitly.
Reporting standards
Case reports follow CARE, which covers the timeline, the diagnostic reasoning, the intervention and the follow-up, and requires patient consent for publication.
Descriptive studies of populations follow STROBE. Prevalence estimates carry confidence intervals; a proportion from a small series is far less precise than it looks.
In practice
Fix, flag or query.
The distinction that keeps an edit defensible: what an editor may change silently, what must be shown to the author, and what only the author can answer.
- Never: attribute improvement to treatment without a comparison group.
- State: whether cases were consecutive or selected.
- Obtain and report: patient consent for case reports.
Reporting standards
- CARE guidelines, for case reports
- STROBE Statement, for descriptive observational studies
Reporting guidelines named here are hosted by the EQUATOR Network. Writing to the checklist for your design before submission is the cheapest available improvement to a methods section.
Related