How We Research & Source Content

A practical explanation of how Off the Couch finds, evaluates and communicates evidence — and how we handle uncertainty when the research is not clear-cut.

Our starting point

Psychology is a research field, but it is also a field where context matters. A single study, statistic or expert quote rarely tells the whole story. Our goal is to build articles from a wider evidence picture rather than treating one source as the final answer.

We also try to avoid a common problem in online mental-health content: presenting emerging ideas, social-media language or broad generalisations as though they are settled clinical facts.

Our research process

Define the question

Before searching, we identify what the article is actually trying to establish: a definition, prevalence estimate, mechanism, risk factor, treatment question, social pattern or practical implication.

Look for the strongest relevant evidence

Where available, we prioritise original research, systematic reviews, recognised clinical or public-health guidance, official records and established professional bodies.

Check context and limits

We consider who was studied, where the research was conducted, how the outcome was measured and whether the evidence supports the broader claim being made.

Compare, rather than cherry-pick

When credible studies disagree, the article should reflect the disagreement rather than selecting only the result that creates the strongest headline.

Separate evidence from interpretation

Research findings, professional interpretation and editorial synthesis are not the same thing. We aim to signal when a conclusion is interpretive rather than directly demonstrated by a source.

Write for people, not just papers

Technical evidence should be translated into clear language without stripping away the caveats that make the evidence responsible.

Source hierarchy

No source type is automatically perfect, but different sources are useful for different questions. As a general editorial approach, we prefer sources that let readers inspect the original evidence and understand how a conclusion was reached.

Primary research & systematic reviews

Useful for understanding specific findings, patterns across studies and the strength or limits of evidence.

Clinical, professional & public-health guidance

Useful when explaining recognised definitions, care pathways, safety guidance and current professional consensus.

Official records & public data

Useful for policy, population data, institutional processes and documented public facts.

Credible secondary reporting

Useful for context and explanation, particularly when it points back to accessible original material. It should not make weak underlying evidence appear stronger than it is.

What we avoid

Using one small study to support a sweeping claim.
Presenting correlation as proof of cause.
Using prevalence figures without enough context about population, place or time period.
Diagnosing public figures or people described in news stories from a distance.
Relying on social-media popularity as evidence that a psychological claim is clinically valid.
Quoting experts in a way that removes important limits or qualifications from what they actually said.
Using statistics without checking what was actually measured.

How we communicate uncertainty

Not every psychology question has a settled answer. When evidence is preliminary, contested, based on limited populations or difficult to generalise, we aim to say so directly.

Words such as “may,” “is associated with,” “research suggests,” or “evidence is mixed” are not weaknesses when they accurately reflect the research. They are part of responsible communication.

AI and editorial responsibility

Tools may assist with research organisation, drafting or editorial workflows, but they do not replace source checking, professional judgment or human responsibility for what is published. Claims that matter should be checked against the underlying evidence rather than accepted because a tool produced confident wording.

If you think we got something wrong

We welcome specific corrections, stronger sources and missing context. Please use the Contact page and include enough detail for us to examine the issue. Our approach to substantive corrections is described in the Corrections Policy.