You found a source that looks good. Before you cite it, you need a fast, reliable way to decide whether it actually holds up. That's what the CRAAP test was designed for — a five-criterion checklist developed by librarians at California State University, Chico that turns source evaluation from gut feeling into a repeatable process. Learn it once and you can vet any source in under five minutes.
PaperDraft is a writing assistant, not a paper generator — the draft is your starting point, not your submission. You are responsible for editing, verifying sources, and following your school's academic integrity policy.
Below, what each letter in the CRAAP test stands for, how to apply it to real sources, and where the framework needs judgment calls rather than checkbox answers.
What the CRAAP test actually is
CRAAP is an acronym for five questions you ask about every source:
- Currency — when was it published, and does that matter?
- Relevance — does it address your specific argument?
- Authority — who wrote it, and are they qualified?
- Accuracy — is the content verifiable against other sources?
- Purpose — why does this source exist, and who benefits?
A source doesn't have to score perfectly on all five. It has to score appropriately for how you plan to use it. A 1980 paper is fine for historical context, wrong for current epidemiology.
Applying CRAAP step by step
Currency: is it recent enough?
Check the publication date. Then ask whether your topic is in a fast-moving field. In AI research, a 2021 paper is already outdated. In medieval history, a 1985 paper can still be canonical.
Rules of thumb:
- STEM and applied fields: default to the last 5–7 years, with older work only for foundational references.
- Humanities: age is usually fine; canonical texts don't expire.
- Social sciences: the last 10–15 years for most claims, older for theoretical grounding.
Check whether the paper has been updated, retracted, or superseded. Retraction Watch and the journal's website both list retractions.
Relevance: is it actually about your topic?
This one sounds obvious and is the one students fail most. A paper being on your general topic isn't enough. Read the abstract and ask:
- Does it address the specific question my paper asks?
- Is the population, timeframe, or context similar enough to mine?
- Can I actually use this in my argument, or am I keeping it because I found it?
If you can't write one sentence explaining how the source supports your thesis, the source isn't relevant yet. Either clarify the connection or cut it.
Authority: who wrote it?
Find the author's credentials and institutional affiliation. Search their name and field — do they have related publications? A track record?
Then check the publisher. Is the journal indexed in Scopus, DOAJ, or Web of Science? Is the book published by a university press or a reputable academic publisher? Anonymous or unattributed content rarely passes authority.
Note: authority isn't just about credentials. A credentialed expert writing outside their field is still outside their field. A neuroscientist writing about economics is guessing, same as you.
Accuracy: does the content check out?
The strongest accuracy test: can you verify at least one factual claim from the source against an independent source?
- Click through the citations. Do they actually say what this source claims they say?
- Check numerical claims against primary data.
- Look for methodological red flags — tiny sample sizes, no control group, conflicts of interest.
Predatory journals fail accuracy checks fast because their cited sources often don't exist or say different things. Five minutes of cross-checking catches most fabrication.
Purpose: why does this exist?
Every source has a purpose. Identifying it isn't about rejecting advocacy — it's about knowing how to weight the source.
- A peer-reviewed journal article: designed to advance knowledge, reviewed by peers.
- A think tank report: designed to influence policy, reviewed internally.
- A corporate whitepaper: designed to sell something.
- A news article: designed to inform a general audience, often on deadline.
Ask: who paid for this, who benefits if it's believed, and would the conclusions be different under different funding? Fund-source disclosures are often at the end — read them.
Worked example: a real source
Suppose you found a 2019 paper on "Social Media Use and Teen Mental Health" in the journal Cyberpsychology, Behavior, and Social Networking.
- Currency: 2019, within the five-year window for social sciences. Acceptable, but check for more recent replications.
- Relevance: if your paper is on adolescent mental health and digital media, this is directly relevant. If your paper is about adult anxiety, this is tangential.
- Authority: authors at known psychology departments. Journal is indexed in Scopus and PubMed. Passes.
- Accuracy: check the cited studies in the literature review. Spot-check one statistic against the original source.
- Purpose: peer-reviewed academic research. Funding source should be disclosed — check for conflicts.
Verdict: probably citable, pending the accuracy spot-check and relevance match to your specific thesis.
Six PDFs vetted and still no draft on the page? PaperDraft gives you a structured first draft — thesis stub, outline, cited opening sections — so you can spend your time evaluating sources instead of starting from zero. It's a drafting assistant, not a submission. Try PaperDraft — free
Common mistakes
Treating CRAAP as a checkbox instead of a judgment tool. The point isn't to score 5/5. It's to know which criterion matters most for this source and this paper.
Skipping the accuracy spot-check. Checking authority and currency is fast. Verifying that cited sources actually say what the paper claims takes longer — and it's the check that catches the most issues.
Applying currency uniformly. A foundational philosophy text from 1975 is fine. A 1975 climate science paper is not. Currency depends on the rate of change in the field.
Ignoring purpose because the source looks neutral. Every source has a purpose. "Academic research" is a purpose too — a tenure-motivated one. Being skeptical of funding isn't cynicism, it's methodology.
Using CRAAP instead of, not alongside, broader credibility judgment. CRAAP handles individual sources. For the bigger question of what makes a source trustworthy at all, read our credible vs non-credible sources breakdown.
How a drafting assistant fits
Once you've run the CRAAP test on your sources and kept only the ones that pass, PaperDraft can take your vetted source list and scaffold a structured first draft — a thesis stub, an outline, and opening sections with citation stubs pointing to the sources you evaluated. You still read each paper, verify the citations, and write the analysis. The tool handles the scaffolding; you handle the judgment, because source evaluation is the part nobody else can do for you.
FAQ
Where did the CRAAP test come from?
It was developed by Sarah Blakeslee and colleagues at California State University, Chico's Meriam Library, and has become one of the most widely taught source-evaluation frameworks in academic libraries.
Is the CRAAP test outdated for digital sources?
The five criteria still apply, but currency and purpose carry more weight for digital-native content. Newer frameworks like SIFT (Stop, Investigate, Find, Trace) complement CRAAP for fast-moving web content.
How long should the CRAAP test take per source?
Five minutes per source once you're practiced. The accuracy spot-check is the time sink — skip it at your own risk.
Can I use CRAAP on books?
Yes. Currency for books is usually edition-sensitive. Check the latest edition, and for reissued classics, check the introduction for scholarly update notes.
What if a source fails one criterion but I still want to use it?
You can, if you frame it honestly. A biased source can be cited as evidence of the bias. An older source can be cited for historical context. The failure becomes a note, not a reason to reject. For the full research workflow, see our guide to writing a research paper or pull the annotated bibliography template to record your CRAAP assessments.