Evidence synthesis · Lesson 1 of 3

The kinds of review, and what each is for

Pick the review your question and your calendar can both support, and name it accurately.

8 min read · Free — no account needed to read

Narrative review reads widely and tells you what the field thinks, with the author's judgement doing the work. Valuable for orientation and for argument; its selection is not reproducible, so it cannot settle a contested question.

Scoping review maps a field systematically: what has been studied, with what methods, where the gaps are. It does not judge quality or pool results. It is the right first move when you do not yet know the shape of a literature.

Systematic review answers a specific question with a pre-specified method: criteria set in advance, a documented search, screening by two people, quality assessed, findings synthesised. It is reproducible, and it takes months.

Meta-analysis statistically combines results from studies measuring the same thing, producing a pooled estimate with more precision than any single study. It requires comparability — combining studies that measured different things precisely produces a precise number about nothing.

Meta-synthesis is the qualitative counterpart, integrating findings across qualitative studies to build a higher-order interpretation. It does not average anything.

Rapid review applies systematic methods with deliberate shortcuts for a decision that cannot wait, and says which shortcuts were taken. Umbrella review reviews the reviews, when a field has many.

Choose by the question and by what you can actually complete. A well-executed scoping review is worth far more than an abandoned systematic one.

See it happen

You need to know within three weeks whether a workplace intervention has evidence behind it.

The read that arrives on its own

Do a systematic review — it is the rigorous option and the decision matters.

The same situation, with this idea

You check what a systematic review takes: typically several months with two screeners. In three weeks you would produce something with the name and none of the method, which is worse than an honest alternative. A rapid review — systematic search, single screener, quality noted rather than formally assessed, shortcuts stated — is deliverable and defensible. Or an existing systematic review may already exist; searching for one is twenty minutes.

After this, you can

match the review type to the time you actually have, and say which shortcuts you took.

The takeaway

Pick the review your question and your calendar can both support, and name it accurately.

Easy to take the wrong way

That systematic is always better. A scoping review answers "what is out there" far better, and a systematic review of a literature that has not been mapped often asks the wrong question.

Try this

Write which review type your question needs and how long that type actually takes in your field.