Two terms that sound like synonyms and aren’t. Students who conflate them end up with a bibliography that’s formally spotless and gets red marks anyway, because half the entries are “citable but not worth citing.” This post pulls the two apart, walks through the hard cases, and gives you a checklist you can run before you add a source to your reference list.
Citable: the formal bar
A source is citable when it can be uniquely identified and retrieved by someone else. That’s it. The question is form, not content.
In practice, that means clear information on:
- Author or editor (person or organisation),
- Title of the work or contribution,
- Year of publication (or at least a defensible date, e.g. “last updated”),
- Publication context in the broadest sense: publisher, journal with volume and issue, stable URL, DOI, or repository.
An Instagram reel with no author and no date isn’t citable. An anonymous PDF of lecture notes making the rounds in your cohort isn’t citable. A ChatGPT reply with no saved chat link and no timestamp isn’t citable either. In those cases you literally can’t produce a reference entry someone could follow — not because the content is bad but because the metadata isn’t there.
Citability is the minimum condition: if a source fails it, there’s no point debating the content. The reference entry couldn’t be built.
Worth citing: the content bar
A source is worth citing when the content is robust enough to support the claim you’re attaching to it. Here, metadata is beside the point. What matters is what’s actually in the text.
Four questions decide whether a source clears this bar:
- Has the content been quality-controlled? Peer review, publisher editing, an editor at a recognised series — some form of external vetting.
- Is the method transparent? Does the text say how the data were collected, with which limits, against what theoretical background?
- Is the source current enough for your topic? In fast-moving fields (AI, medicine, law) a five-year-old study is often out of date. In historical or theoretical disciplines, foundational texts from the 1960s are an obvious yes.
- Does the author have standing in the field? Have they published there, do they teach there, are they cited? Or is this an opinion piece with no track record on the topic?
A source doesn’t need four yeses. But every “no” is a reason to ask whether you really need this source or whether a better one exists.
The four-quadrant matrix
Citability and worth-citing are independent checks. That gives four cases:
| worth citing | not worth citing | |
|---|---|---|
| citable | The normal case: peer-reviewed journal articles, monographs from academic publishers, relevant standards. Cite. | Newspaper opinion piece, corporate white paper, many textbooks outside their field. Only cite if the claim really sits there and no better source exists. |
| not citable | Smart content, formally unreachable: anonymous lecture notes, oral remarks in a seminar, undated social media posts. Don’t cite. If the content matters, chase the original. | A TikTok rant with no evidence. Forget it. |
The interesting cell is top-right: citable but not worth citing. That’s where almost every supervisor argument happens. The form is clean, the quality isn’t, and the judgement is a judgement call.
Edge cases students keep running into
Wikipedia
The edge case where both bars actually fail under the stricter house rule: a specific version is formally retrievable via its permalink and access date, but in most fields Wikipedia counts as not citable at all because no author stands behind the text as a person. It’s rarely worth citing either — collectively edited, seldom peer-reviewed, potentially shifted between your access and your examiner’s. Use Wikipedia to orient yourself and chase down the references at the bottom — then cite those directly. The exceptions (Wikipedia as research object) are in our Wikipedia as a source post.
Quality newspapers (NYT, Guardian, FT, NZZ)
Citable: yes. Worth citing: situational. For current events, legislative moves, political developments a quality paper is a legitimate source, including in theses. For scientific claims (“studies show that…”) you go back to the study itself, not to the newspaper article covering it.
Corporate white papers and consulting reports
Citable: usually yes. Worth citing: handle with care. A McKinsey or Deloitte report is a marketing product that promotes the publisher’s service. For a market or industry overview in your introduction you can use it; as evidence for a substantive thesis, pair it with a second, independent source.
YouTube lectures by professors
Citable: if title, channel, date and a stable link are all there. Worth citing: often yes, but only when the content isn’t also in a publication by the same author. In that case you cite the publication. A YouTube lecture is a weaker version of the paper — unless the clip really contains a point that isn’t anywhere else.
Student theses
Bachelor theses: generally not worth citing. Not peer-reviewed, often not publicly accessible, small samples. Master theses: only if published and directly relevant. Doctoral theses: yes — they were examined.
Researcher blogs
Citable: yes, with a permalink and date. Worth citing: only for positions and interpretations, never for empirical evidence. If the blogger is summarising their own study, cite the study — not the blog.
Standards (DIN, ISO, ANSI)
Citable and worth citing: yes, in technical and quality-related fields often non-negotiable. For how to format the reference, see Citing standards (DIN/ISO).
ChatGPT and other LLMs
Citable: only with a saved chat link or prompt documentation plus a date — and even then contested. Worth citing: almost never, because the reply isn’t reproducible and has no verifiable source behind it. See ChatGPT as a source.
What to do when a source doesn’t clear the bar
When you catch a source that’s citable but not worth citing, you have three options:
- Trade up to the primary source. Most weak sources themselves cite a better one. Follow the footnote, read the original — and cite that.
- Add a second, stronger source. If the weak source makes the point but a peer-reviewed one says the same thing, you can cite both. The stronger one carries the claim.
- Soften or drop the claim. If no worth-citing source supports it, the claim is probably less safe than you thought. Reframe it as a conjecture or cut it.
What you shouldn’t do: dress a weak source up in impeccable formatting and hope the form hides the content gap. Examiners see through that at a glance.
Checklist before you add a source
Before a source lands in your bibliography, walk through:
Citable:
- Author or editor is identified?
- Title is unambiguous?
- Year is stated (or a defensible date)?
- Publication context works (publisher, journal with volume/issue, DOI, stable URL)?
Worth citing:
- Was there external quality control (peer review, publisher editing)?
- Is the method disclosed?
- Is the source current enough for your field?
- Does the author have standing on this topic?
Eight checks. Eight ticks and you’re clear. One blank and you think again before you type.
Both checks, before your examiner runs them
Both checks stay on you. Supervisors read your thesis, not your sources — they can tell when a reference sounds weak, but they rarely open every PDF to compare the claim against the page.
That’s where Acurio comes in. You upload your thesis along with the cited PDFs, and Acurio checks every reference against the source itself. Its job is the piece that style guides and plagiarism scanners can’t do — whether the page number is right and whether the claim really sits on the page you cited.
Citability and worth-citing you decide yourself, with the checklist above. The third step — whether the reference and the source actually match in content — Acurio takes off your plate before the second examiner does.
Two terms, two checks, both belong in your workflow. Students who run both end up with fewer sources but ones that hold up. In any thesis that beats a long reference list that falls apart on inspection.