Does it exist, and can you reach it
Retrieve it before anything else, because this test fails fastest and disqualifies a source outright. Search the exact title in your library and check that the identifier resolves. A source that will not come back for you will not come back for your marker either.
This check has become non-optional rather than cautious. Generated reference lists produce entries with real journals, plausible authors and page ranges that lead nowhere, formatted immaculately, and no software will tell you. A minute per entry settles it.
Who produced it, and what for
Authorship tells you about expertise; the publisher tells you about incentive. A professional body publishing guidance, a government agency publishing statistics and a manufacturer publishing a white paper on its own product are three different kinds of source, and only one of those is trying to sell you something.
None of that makes commercial or advocacy material unusable. It makes it usable for some claims and not others, which is a question about placement rather than about quality.
How old is it, relative to the claim
Age matters differently depending on what the source is doing. A foundational theoretical work does not expire. A prevalence figure, a guideline or a statement about current practice does, sometimes within a couple of years.
So the question is never simply how old, it is whether anything has changed since. For clinical guidance that check is quick, because issuing bodies publish revision dates and mark superseded documents.
Does it actually support your sentence
The last test is the one people skip and it fails more sources than the first three combined. Read the specific passage you intend to cite, not the abstract, and compare it against the sentence you have written.
Abstracts report findings without the conditions attached to them, which is how a result that held in one population under particular circumstances ends up cited as though it held generally. Where the source is narrower than your sentence, narrow the sentence. That is almost always the better fix, and it is usually the more accurate claim as well.
Run them in that order
The order matters because each test is cheaper than the one after it. Retrieval takes seconds and disqualifies a source outright. Provenance takes a minute. Currency takes a minute more. Reading the cited passage properly takes the longest and is the one worth spending real time on.
Applied in that sequence a bad source costs you a minute. Applied in reverse you can spend twenty minutes carefully reading something that was never going to survive the first check, which is how source work comes to feel so much longer than it actually is.
Why this is worth the habit
A reference list is the part of your work a reader can verify without taking anything on trust, which makes it the fastest available signal about how carefully the rest was done. A marker who opens two sources and finds both exactly as described stops checking.
One who opens two and finds a dead link and a mismatch starts reading everything differently. That shift costs far more than the individual errors do, and it happens silently, which is why the four tests are worth running on every list you submit.
Questions this raises.
Is a source with no author unusable?
Not necessarily. Organisational authorship is normal and legitimate for guidance, standards and statistics, where the issuing body is the author. What should make you cautious is a source with no identifiable author and no identifiable organisation behind it, because there is then nobody whose expertise or incentive you can assess.
Do textbooks count?
For definitions and background, usually yes, and your rubric may restrict them. What a textbook cannot do is serve as evidence of a current effect, because it is a summary written some time before publication. Where a textbook cites a study you want to rely on, find that study and cite it directly.
What about a source behind a paywall?
If you can reach it through your library it is fine, and if you cannot then it fails the first test regardless of quality. Interlibrary loan is worth using where there is time. What is not acceptable is citing something you have only read the abstract of, which is what usually happens instead.
Should I check every single reference?
Every one, and it goes quicker than it sounds because the first test disposes of most problems in seconds. On a list of twenty-five references the whole exercise runs to roughly half an hour. Set against a marker finding a source that will not open, that is among the best-value time in academic writing.
Does a source have to be peer-reviewed?
It depends on the claim. Peer review is a quality signal for research findings and it is irrelevant for a statistic published by a government agency or a standard issued by a professional body, neither of which goes through it. Ask what the source is being used for, then ask what would make that particular use trustworthy.
Clara Fairbairn
MA, Writing, on the essays and feedback side of the hall. Writes for The Footnote and coaches the students who bring this work to the desk. The rest of the specialists.
Elsewhere in the hall.
Send the assignment and the deadline.
A specialist credentialed in your own field reads what you actually have and replies within two hours with a plan and a price. The reading costs nothing, and if the honest answer is that we are not the right hall for it, you will get that instead.
