Systematic, scoping and meta-analysis reviews

"Systematic review" gets used for any review that felt thorough. It has a specific meaning, and using the term loosely in a dissertation invites a marker to hold you to the full standard.

The four types

Narrative review

The standard dissertation literature review. You search purposefully, organise thematically and build towards a gap, but you do not pre-register a protocol or appraise every study formally.

Perfectly legitimate, and right for most undergraduate and many master's dissertations. Just do not call it systematic. Ourliterature review guide covers the method.

Scoping review

Used when you want to map a field rather than answer a narrow question. Appropriate where the evidence is too new, too varied, or too methodologically mixed for a systematic review to work.

Typical questions:

• What research exists on digital mental health interventions for UK adolescents?
• What outcome measures have been used in studies of school attendance?
• Where are the gaps in the literature on carer wellbeing?

A scoping review still uses a documented search strategy and explicit inclusion criteria. What it normally omits is quality appraisal, because the aim is coverage rather than synthesis of the best evidence. The Arksey and O'Malley framework, refined by the JBI, is the usual reference.

Systematic review

A narrow question, answered by a protocol specified before searching begins. The defining feature is reproducibility: another researcher following your protocol should retrieve the same studies.

The stages

1. Define the question, usually with a framework such as PICO
2. Write the protocol before searching, and register it where required
3. Search systematically across named databases with recorded terms. Google Scholar alone is not reproducible enough, as our Google Scholar guide explains
4. Screen titles and abstracts, then full texts, against stated criteria
5. Appraise quality using a recognised tool such as CASP or a Cochrane risk-of-bias instrument
6. Extract data into a standard form
7. Synthesise, narratively or statistically
8. Report against PRISMA

The student caveat

A full systematic review uses two independent reviewers screening in parallel, with a third resolving disagreement. A dissertation student working alone cannot do that.

This is not a problem provided you say so. Describe your work accurately as a "systematic-style review" or "systematic review conducted by a single reviewer", and name single-reviewer screening in your limitations. Markers penalise the overclaim, not the constraint.

PRISMA

PRISMA is a reporting standard, not a method. It tells you what a review must disclose so a reader can judge it. Two components matter for a dissertation.

The checklist

Twenty-seven items covering everything from the title to the funding statement. Work through it before submission and you will find the two or three things you forgot to report.

The flow diagram

The four-stage figure showing what happened to every record. Markers look for this immediately.

Every number must reconcile. Exclusions at full-text stage need reasons; exclusions at title stage do not.

Meta-analysis

A statistical layer on top of a systematic review, pooling effect sizes from multiple studies into a single estimate with a confidence interval.

It requires all of the following:

• Enough quantitative studies, generally a minimum of five
• Comparable outcome measures across those studies
• Reported statistics sufficient to calculate effect sizes
• Acceptable heterogeneity, usually assessed with I²
• Software: RevMan, R with the metafor package, or Stata

Where studies measure different things in different ways, pooling them produces a precise number that means nothing. Narrative synthesis is the honest alternative and markers respect the judgement.

Choosing

Choose on the question, then check feasibility, then describe accurately what you actually did. The commonest error in student review dissertations is not method choice but overclaiming the label.

Last updated: 30 July 2026 · Written by the AssignWise editorial team ·Spotted something wrong?