Key Takeaways — Chapter 15: Literature Reviews and Research Synthesis

The summary card. If you remember nothing else, remember this page.

The one idea

A literature review makes an argument; it doesn't catalog. Its job is to use many sources, collectively, to build toward a claim about what's known and the gap your work will fill — not to report each source in turn. A pile of accurate summaries is a list, not a review.

🚪 Threshold concept: The sources are your evidence, not your subject matter. The grammatical tell is diagnostic: in a summary, sources are the subjects of your sentences ("Nakamura found…"); in a synthesis, ideas are the subjects ("The predictors are structural…"), with sources in parentheses as support. When your sentences start with author names, you've slipped back into cataloging.

Synthesis vs. summary — the distinction the whole chapter turns on

  • Summary = restating one source. Source-centric. Necessary, but a stack of them isn't a review.
  • Synthesis = combining many sources to make a point none states alone. Idea-centric, sources as evidence. This is what a review is for.
  • The fastest diagnostic: what's the grammatical subject of each sentence — an author, or an idea? And: could you shuffle the paragraphs into any order and lose nothing? (If yes, it's a list.)

The synthesis matrix — your core tool

A table: sources down the rows, themes across the columns, your own-words notes in the cells. Let the themes emerge from the sources. - Read down each column ("what does the whole literature say about this theme?") → that's synthesis. - Read the empty cells → that's the gap. A sparse column is a hole in the field, not your reading. - The matrix is also your outline: each populated column becomes a thematic section. It moves the thinking earlier, where it's manageable.

Organize by theme, not by source

Each section pulls together every source on one theme and states how they relate. Open paragraphs with a claim, then bring sources in as evidence. Connective verbs do the synthesizing: converge, in contrast, building on, what remains unexamined. (The given-new contract from Ch 8 keeps it flowing.) The source-by-source draft is fixed by reorganizing, never by editing sentences one at a time.

Finding and naming the gap

  • A real gap = an unanswered question, a contradiction, an untested context with a reason it matters, a methodological limit, or an unconnected synthesis. Test every candidate with "so what?"
  • A false gap = "nobody has studied my exact niche." Novelty ≠ significance; that's true of infinitely many topics.
  • Honesty: search specifically for the paper that would close your gap. If it exists, engage it — don't conceal it. A concealed gap detonates in review.

Review types — don't claim a method you didn't use

Type What it is
Narrative Judgment-selected, thematic, argumentative — the usual case (paper intros, theses)
Systematic Exhaustive, pre-registered, reproducible search — when the review is the contribution
Scoping Documented search to map a field's extent and gaps

Calling a judgment-selected review "systematic" is a misrepresentation a reviewer exposes by asking for your search protocol.

The reader's only question is "how is yours different?" Group prior work into 2–4 categories, characterize each by its limitation, and end by positioning your contribution on the axis no one else covers. Cite in groups ("[1], [3], [5]"); be fair to the work you distinguish yourself from (critique scope, not competence).

Managing many sources (Ch 11's tools, scaled)

Capture every source in Zotero immediately; annotate in your own words (patchwriting insurance and a ready-made matrix cell); build the matrix incrementally; stop at saturation (new papers add no new themes).

The one-line test for any review paragraph: Is the subject an idea or an author? Could I shuffle the sentences and lose nothing? Does it lead to a gap? If the answers are "author," "no loss," and "no gap," you've summarized — reorganize by theme and lead with a claim.