Network-based Mapping Review

A plain mapping review catalogues studies by topic, population, intervention, or outcome category and presents findings in tabular or narrative form. A network-based variant adds a relational layer: it constructs a citation, co-authorship, or co-word network and uses community detection to reveal which papers, authors, or themes cluster together structurally. The result shows not just what exists but how the evidence base is organised internally.

Each reveals a different facet of the field. Citation and co-citation networks show intellectual lineage and shared theoretical foundations. Bibliographic coupling networks reveal papers that draw on similar bodies of prior work, making them useful for grouping current research directions. Co-authorship networks map collaboration structures. Co-word and keyword co-occurrence networks map conceptual proximity. Most network-based mapping reviews construct at least two types and compare them to obtain a richer picture.

As a practical threshold, at least 100 records are needed to produce a network with interpretable cluster structure. Networks built from fewer than 100 documents tend to produce trivially few or poorly differentiated clusters. For robust field mapping, 500 or more records are ideal. If your literature is small, a scoping review with a traditional thematic synthesis is more appropriate.

Not in the same way as a systematic review. The purpose of a mapping review is breadth — to chart what exists — rather than to synthesise effect estimates from high-quality studies alone. Quality appraisal is therefore typically not performed, though the review should acknowledge that network prominence (high citation counts) does not equal methodological quality. If effectiveness conclusions are needed, the map should feed into a subsequent targeted systematic review.

VOSviewer is the most widely used tool for bibliometric network visualisation and requires no programming. The R bibliometrix package offers greater analytical flexibility and reproducibility. Gephi provides more advanced network visualisation and layout options but has a steeper learning curve. CiteSpace is popular in certain disciplines (information science, environmental science) for temporal co-citation analysis. All are free and actively maintained.