Most literature reviews read like a stack of summaries stapled together: "Smith (2019) found X. Jones (2021) found Y. Lee (2022) found Z." That's not a literature review — it's an annotated bibliography with the joints still showing. A literature review is an argument about what the existing research actually tells us, including where it disagrees with itself, not a list of papers you happened to read.
It's called the "source dump": one paragraph per paper, in roughly the order you found them. It feels thorough while you're writing it, because you're clearly covering a lot of ground. But it can't show how the studies relate to each other, because each one gets sealed off in its own paragraph. Markers aren't grading how many papers you read — they're grading whether you can show what the body of evidence, taken together, actually adds up to.
Group your sources around the ideas, variables or debates they actually address, not the order you found them in. If three papers all bear on "how severity affects treatment response," they belong in the same section even if they were published a decade apart and never cite each other — because your review is making a point about severity, and all three are evidence for or against it. Chronological ordering only earns its place when tracing how a field's thinking changed over time is specifically the point of the section.
A genuine review names the disagreement instead of smoothing over it: "While the larger cohort study and the recent meta-analysis both report a moderate effect, the single RCT in this area found none — a gap usually attributed to differences in how severely affected each sample was." That sentence is doing real analytical work. A list of individual findings, however accurate each one is, never does.
The honest takeaway: a literature review that only proves you read things will always read worse than one that argues something. Organise around the debate, not the reading list.