Why it matters when you are writing
The single most common failure is the annotated-bibliography structure: a paragraph per study, in the order you read them, each summarizing one paper. It demonstrates reading and nothing else. A synthesis instead makes claims about the literature — that two camps disagree about mechanism, that every study of a phenomenon has used one population, that a finding replicated in labs has never been tested in the field — and cites several sources per claim.
A practical route there is a synthesis matrix: rows for sources, columns for the dimensions you care about — sample, method, construct, finding, limitation. Filling it in surfaces the patterns that become your paragraph topics, and the empty cells are frequently the gap you have been trying to articulate. It also stops you from writing about a study you have only skimmed, because the row exposes what you did not extract.
The review has to end somewhere specific. A supervisor reading a draft is checking whether the closing paragraph makes your research question feel inevitable — whether the gap you name is the one the preceding pages actually established. If the question could be swapped for a different one without changing the review, the review has not done its job yet.
