English Composition › Section N · The annotated bibliography › Topic N.1
Compiling and evaluating sources
With a library and the internet to hand, you will never run short of sources; the challenge is deciding which ones deserve your readers' trust. Evaluating sources well means asking who made them, for what purpose, and what other sources say about them.
Compile widely, then judge
At the start of a project, gather generously: books, journal articles, reports, websites, periodicals and documents that might hold useful ideas. You will not use them all. As your thesis takes shape, evaluate each source and keep those that support your claim, challenge it, or help you answer counterarguments. A working bibliography, your list of sources with their details, keeps this organised; a finished bibliography shows readers your research is trustworthy and credits the authors you draw on. The earlier lesson on types of evidence explained primary, secondary and tertiary sources; this lesson asks how far to trust each one.
A hierarchy of authority
Sources fall roughly into four tiers. First, peer-reviewed academic publications: journal articles and scholarly books evaluated by experts before publication. Second, credible nonacademic sources: reports by public bodies, whether national governments or the World Health Organization, long articles from well-regarded news organisations, and books by credentialed experts. Third, short articles from credible websites and periodicals, useful for facts and for leads to fuller sources. Fourth, sources with clear bias or unclear authority. These cannot support a claim on their own, but they can show you the positions in a debate, suggest keywords and point towards better sources.
Peer review and the academic article
Peer review means that other experts in the field, usually anonymously, evaluated a work before it was published. It is a strong quality signal, but not a guarantee: reviewed studies can contain errors, and a single study can be overturned by later ones, so look for findings that several studies confirm. Academic articles, especially in the sciences and social sciences, usually follow a pattern: an abstract summarising purpose, methods and findings; an introduction posing the question; a review of earlier research; data and methods; results; a conclusion or discussion; and a reference list. Knowing the pattern lets you find what you need quickly.
Value depends on purpose
A source is not simply good or bad: its value depends on what you need it for. A magazine's one-sided account of a protest is poor evidence of what happened. But if you are analysing how media coverage shaped public opinion of the protest, that same account becomes valuable evidence. So ask what job you want the source to do, whether it can do that job, and whether it might serve better in another role. Also weigh time and perspective. An old statistic misleads if you describe today, but is ideal as a baseline for change over time. And every source reflects its creator's assumptions, evidence and reasoning.
The CRAAP test
The CRAAP test, a checklist devised by librarians at California State University, Chico, gives five sets of questions. Currency: when was it published or updated, and does your topic need recent information? Relevance: does it relate to your question, and who is its intended audience? Authority: who is the author or publisher, and what are their credentials? Accuracy: where does the information come from, is the evidence verifiable, and has it been reviewed? Purpose: is it meant to inform, persuade, sell or entertain, and are its biases acknowledged? A source can pass some tests and fail others, so record the result for each letter rather than a single verdict.
Worked example: ranking sources for one question
Question: "Do school gardens improve students' eating habits?" Four invented sources, offered as examples: (A) a peer-reviewed study of 30 schools in a nutrition journal; (B) a national health agency's report on child nutrition; (C) a 400-word local news story about one school's garden; (D) a seed company's blog post claiming gardens "transform" diets. Analysis: 1. A is first tier: reviewed, with stated methods. 2. B is second tier: credible and broad, with national data. 3. C is third tier: a vivid example, not proof of a pattern. 4. D has a commercial purpose, so it fails the purpose test; use it only for keywords or leads. 5. Together, A and B carry the argument and C illustrates it.
Bias, agendas and domains
Every source has a perspective, but some also have an agenda or a conflict of interest, such as a study of sugary drinks funded by a drinks company, or a think tank with a strong political mission. Bias does not automatically make a claim false; it means you should check the claim against independent sources. Web addresses give clues, not verdicts. A .gov address marks a US government body and .edu a US college or university, though student and personal pages can sit there too. Anyone can register a .org address, and many belong to advocacy groups that argue for a cause. Look for an "About" page and author credentials, and prefer writing that presents evidence neutrally.
Lateral reading and SIFT
Researchers at Stanford who studied professional fact-checkers found that they rarely judged a website by studying the site itself, which can look polished while being unreliable. Instead they read laterally: they left the page and opened new tabs to see what other sources said about the site and its authors. The educator Mike Caulfield's SIFT method packages this habit: Stop before you use or share a source; Investigate the source; Find better coverage of the claim; Trace claims, quotations and images to their original context. These habits matter more as manipulated images and AI-generated text become harder to spot. Misinformation is false information spread regardless of intent; disinformation is spread deliberately.
Worked example: SIFT on a polished website
A student finds "Screen Health Institute", an invented site, claiming that blue-light glasses cut eye strain by 70 percent. 1. Stop: the figure is striking, so she pauses. 2. Investigate: she searches the institute's name in a new tab and learns it is run by a company that sells the glasses. 3. Find better coverage: she searches the claim and finds a review of several trials in a medical journal reporting little or no benefit. 4. Trace: the site's "study" turns out to be a customer survey on its own website. 5. Decision: she does not use the site as evidence, though it might illustrate how products are marketed.
Working smarter with good sources
Once you have one strong article, use it to find more. Read abstracts first to judge relevance, skim promising articles, and save close reading for later. Mine the article's literature review and reference list for related sources, look up the authors' other work, and use a database's "Cited by" feature to find later work that responds to it. If a publisher's site asks for payment, record the details and check your library first.
Common mistakes to avoid
First, judging a website by how professional it looks. Second, treating peer review as proof that a finding is final. Third, rejecting every biased source, when some are useful as examples of bias. Fourth, citing an outdated statistic to describe the present. Fifth, assuming every .org site is neutral. Sixth, accepting a striking figure without tracing it to its original study. Seventh, keeping only sources that agree with you.
Key takeaways
Compile widely, then evaluate. Authority runs from peer-reviewed academic work, through credible nonacademic reports and books, to short credible pieces and finally sources with clear bias or unclear authority, which serve only as leads. Peer review is a strong signal, not a guarantee. A source's value depends on your purpose, its date and its perspective. Apply the CRAAP test: currency, relevance, authority, accuracy, purpose. Read laterally and use SIFT. Check agendas and domains, and use good sources to find more.
Flashcards
Flashcard
Your growing list of possible sources and their details
Test your understanding
Question 1 of 5
In the hierarchy of authority, where does a long report from the World Health Organization belong?
Question 2 of 5
What does lateral reading involve?
Question 3 of 5
"What are the author's credentials, and who is the publisher?" Which part of the CRAAP test asks this?
Question 4 of 5
A student finds a one-sided magazine account of a historical strike. When would it be most useful?
Question 5 of 5
What is the most accurate statement about peer review?
