How to Evaluate Academic Sources: The CRAAP Test

Not all academic sources are equal in credibility and reliability. A peer-reviewed article in a high-impact journal reporting a well-replicated finding is very different from a conference paper reporting a preliminary study or a popular science book summarising research at secondhand. Knowing how to evaluate source quality quickly is as important as knowing how to find sources — and the CRAAP test gives you a systematic framework for doing it.

The CRAAP test: a rapid evaluation framework

The CRAAP test applies five criteria to any source:

Currency — How recent is it?

For most purposes, sources published within the last 10 years are preferred. Some considerations:

• Sciences and medicine: 5 years is often the effective horizon for empirical findings
• Social sciences: 10 years is generally acceptable; foundational works may be older
• Humanities: foundational texts may be decades old and still authoritative
• Rapidly evolving fields (AI, technology policy): even 3-year-old sources may be outdated

Check: Publication year in the citation. Also check if a newer edition exists.

Relevance — Does it address your research question?

A credible source that does not address your specific question is not useful as evidence for your argument. Evaluate:

• Does it directly address your topic, or only tangentially?
• Does it address your population, context, or time period?
• Is it at the appropriate academic level?

A general overview is useful for background but weak as primary evidence for a specific claim.

Authority — Who wrote it and are they credible?

For journal articles:

• Is the author affiliated with a university or research institution?
• What are their credentials in this specific field?
• Have they published widely in this area? (Google Scholar author profile)

For books:

• Is it published by a university press (Oxford, Cambridge, Harvard, MIT, etc.) or a reputable academic publisher (Routledge, Sage, Springer)?
• Does the author have scholarly credentials?

For institutional reports:

• Is the publishing organisation independent and well-established?
• What are their known interests or funding sources?

Accuracy — Is the information supported by evidence?

• Does the source cite its evidence?
• Are claims presented as speculative or as established?
• For empirical work: is the methodology described?
• Has this finding been replicated or challenged elsewhere?

A source that makes strong claims without evidence or citations should be used with caution regardless of the author's credentials.

Purpose — Why was this source written?

Understanding the purpose shapes how you should use the source:

• Peer-reviewed research article: to report findings and contribute to knowledge
• Review article or meta-analysis: to synthesise existing research
• Policy report: to inform or advocate for a policy position — check for funder influence
• Textbook: to introduce a field; generally reliable for established knowledge but not cutting-edge
• Popular science book: to communicate to a broad audience; secondary at best
• Press release: to promote institutional interests — almost never citable

Journal quality indicators

Impact factor

The Impact Factor (IF) measures the average number of citations papers in a journal receive per year. Higher IF generally indicates greater influence, though it varies substantially by discipline.

IF range	Rough interpretation
0–1	Low impact; may be legitimate but small-scale
1–5	Moderate; solid for most disciplines
5–20	High impact; major journals in most fields
20+	Very high; top journals in science/medicine (Nature, Lancet)

Limitation: IF is discipline-specific. An IF of 3 is excellent in humanities, average in social sciences, and low in medicine. Compare within the field.

Indexing

Journals indexed in Web of Science or Scopus have met minimum quality standards. Journals indexed in PubMed/MEDLINE have met standards for life sciences.

If you cannot find the journal in these databases, approach the source with extra caution.

Quartile ranking (Q1–Q4)

SCImago Journal Rank categorises journals in their field by quartile: Q1 is the top 25%, Q4 the bottom 25%. For most university work, Q1 and Q2 journals are the standard to aim for. Check at: www.scimagojr.com

Evaluating specific source types

Peer-reviewed journal articles

Generally the most reliable. Check:

• Publication date
• The journal's peer-review status
• Whether the methodology is described and appropriate
• Whether the sample size is adequate for the claims

Systematic reviews and meta-analyses

Typically the highest quality evidence for empirical questions because they synthesise multiple primary studies. Look for PRISMA-compliant systematic reviews for health/medicine topics.

Books from academic publishers

Reliable for theoretical frameworks, comprehensive treatments, and established knowledge. May lag primary research by several years (publication timelines). Avoid relying heavily on textbooks as primary evidence — they are secondary sources.

Government and institutional reports

Credible for statistics and policy analysis, but note the publishing organisation's interests. UK ONS statistics, US Census data, and IPCC reports are authoritative in their domains. Industry-sponsored research warrants extra scrutiny.

Conference papers

Often preliminary research before journal publication. Can be cited for cutting-edge findings not yet in journals, but lower evidentiary weight than published peer-reviewed articles.

Grey literature (working papers, preprints)

Useful for recent research, but not peer-reviewed. arXiv preprints in CS and physics are generally credible; SSRN papers in economics are widely cited but variable in quality. Note in your citation that the source is a preprint.

Quick source evaluation checklist

Before citing any source, run through these five checks:

• Peer-reviewed or institutional? Is this a peer-reviewed journal article, a publication from a credible institution, or an academic book from a reputable press?
• Recent enough? Is the publication date appropriate for your field and question?
• Relevant? Does this source specifically address my claim, not just the general area?
• Evidence-based? Does the source cite its evidence? Are the claims proportional?
• No known quality issues? Is it indexed in a major database? Is the journal reputable?

For guidance on finding sources, see How to Find Academic Sources. For formatting them correctly once evaluated, use the Citation Reference Formatter.

Frequently asked questions

What is the CRAAP test?

The CRAAP test is a framework for evaluating sources: Currency (how recent is it?), Relevance (does it address your research question?), Authority (who wrote it and are they credible?), Accuracy (is the information supported by evidence?), and Purpose (why was it written — to inform, persuade, sell?). It was developed by Meriam Library at California State University, Chico. It works for evaluating any source type, from academic journals to websites.

How do I know if a journal is reputable?

Reputable journals are indexed in major academic databases (Web of Science, Scopus, PubMed), have a clear peer-review process stated on their website, have been publishing for several years, have an editorial board of recognisable scholars, and have an impact factor (available through Journal Citation Reports). Be cautious of journals with very broad, generic names, journals that email unsolicited invitations to submit, and journals with unusually fast review times (weeks rather than months).

What are predatory journals and how do I avoid them?

Predatory journals are pay-to-publish outlets that claim to be peer-reviewed but conduct no meaningful review. They have proliferated since open-access publishing became common. Warning signs: unsolicited email invitations to submit, promises of fast publication (days or weeks), fees to publish (legitimate open-access journals have fees, but these are disclosed upfront and the journal is indexed), very broad scope ('all fields of science'), no editorial board listed, or listing fictitious scholars on the board. Check the journal against Beall's List (now maintained by independent academics) or verify indexing in Web of Science or Scopus.

Is a high citation count a reliable indicator of source quality?

Citation count indicates influence, not quality. Highly cited papers have been extensively engaged with, which generally means they have shaped the field — but papers can be highly cited for being controversial, for being the first to measure something (regardless of quality), or for historical reasons. A highly cited paper from a major peer-reviewed journal in its field is generally reliable. A highly cited preprint with methodological concerns should be cited with caution.