Triage - which papers to actually read

1 · Learn the move · Comparison matrix

A literature search returns eighty hits; the week has room for five real readings. Triage is not skipping the reading - it is aiming it. The comparison matrix does the aiming: your criteria as columns (system match, method match, sample size as reported in the abstract, recency, relevance to your specific question - not the topic, the question), the hits as rows, and one honest constraint over all of it: the scoring works from titles and abstracts only, and the output must say so. An abstract-scored matrix ranks what deserves your attention; it establishes nothing about what the papers show - methods sections change stories, which is exactly why the top five earn a full read. The model builds the table and holds the honesty line; you pick the five and do the reading.

Triage these search results by comparison matrix. My question, precisely: [the question, not the topic]. Criteria columns: system match to [my system], method match to [methods I can compare against], reported n (from abstract; UNKNOWN if unstated), year, and direct relevance to the question. Rows: the [N] hits pasted below, titles + abstracts. Rules: score ONLY from the pasted text; the header of your output states "scored from titles and abstracts only - not read"; any criterion an abstract does not address = UNKNOWN, never inferred. Rank, recommend 5 for full reading, one line each on why.
[paste hits]