How Many Sources Does a Research Paper Need?

Why one source per page holds as a floor

The rule survives because it matches the paragraph economics of academic prose. Each double-spaced page holds roughly two developed paragraphs; a paragraph that makes a claim needs at least one piece of citable evidence behind it. Not every paragraph is evidentiary — introductions, transitions and conclusions argue from what's already established — which is why the ratio lands at about one new source per page rather than per paragraph.

Rubrics enforce the ratio indirectly. 'Thoroughly researched' papers sit at or above the floor; 'needs more support' comments appear one or two sources below it. The floor also protects you during grading disputes: a paper with 12 sources for 10 pages can absorb a disagreement about one source's quality and still pass. A paper with 5 sources fails the moment any one of them wobbles.

The numbers at a glance

Source counts by assignment length — the range, and where each row actually shows up in coursework. Skim it first; the sections below explain why the range exists and which row applies to you.

The source hierarchy: what fills the count

Peer-reviewed journal articles are the gold standard everywhere, but the acceptable mix shifts by field. Social sciences accept government datasets and major think-tank reports alongside journals. Humanities count scholarly books and edited collections as first-class citizens — sometimes over articles. Engineering and lab sciences weight conference papers and standards documents. Health fields add clinical practice guidelines, which function as top-tier sources even though they're not journal articles.

What never fills the count at any level: Wikipedia and similar wikis, generic content-farm articles, AI-chat outputs, and most commercial websites. Encyclopedias (including scholarly ones) serve as background reading, not cited sources, in most undergraduate rubrics. News articles appear in role — a specific event, a statistic with attribution — but a research paper built on news coverage rather than the studies behind it will draw the 'go to the primary sources' comment every time.

The over-citing trap: when 25 sources lose to 10

More sources is not a linear quality signal. Past roughly 1.5 sources per page, papers enter summary mode: each source gets a sentence or two, analysis thins, and the argument dissolves into 'researchers say X; other researchers say Y'. Graders have a specific comment for this — 'literature list without synthesis' — and it caps papers at B territory regardless of how much work the collection took.

The tell is in your own draft: if you can't state in one sentence what a cited source argues and why it matters to your thesis, you found it, not used it. Strong papers put 30-50 percent of their citations on three or four load-bearing sources that the argument genuinely depends on. Prune before submission: every reference you can't defend in a sentence is a candidate to cut, and cutting usually improves the ratio of analysis to summary that rubrics actually measure.

A source-gathering workflow that fits real deadlines

Start with a review article or handbook chapter on your topic; its bibliography is a pre-vetted map. From it, pull one-third of your target count, favoring recent work. Then run two database searches — one broad (keywords from your thesis), one narrow (the specific population, method or case you're studying) — adding recent items the review missed. This gets a 10-source paper to 12 candidates in 90 minutes.

Triage with the three-question skim: What's the claim? What's the method or evidence? What's the one number or finding you might cite? Two minutes per candidate; anything unclear after that goes back. The survivors — usually 60 percent — get a deeper read with notes taken as paraphrases in your own words, which become draft paragraphs later. Papers built this way cite 10 sources that were actually used, which is the outcome the number was always standing in for.

A five-minute audit of your own source list

Before submitting, run the one-sentence test on every reference: can you state, without opening the PDF, what this source argues and where it supports your paper? Sources that fail the test are collected, not used — cut them or promote them by actually integrating their evidence. Count what remains; that number, not the folder in your reference manager, is what satisfies the requirement.

Two composition checks finish the audit. First, no more than a third of your list should be non-scholarly — news pages, organizational posts, encyclopedias. Second, your three or four load-bearing sources should appear in the text two or more times each; if every source appears exactly once, your paper is a relay of summaries, and the fastest fix is deepening the analysis around your best two sources. The citation generator below formats whatever survives the audit in APA, MLA and Chicago.