The short version. Cite anything that is not your own idea and not common knowledge, cite it at the moment you use it, and give a page number whenever you quote or point at a specific claim. Either APA or Chicago author–date is fine in my courses — pick one, use it consistently, and let Zotero handle the punctuation.
Why we do this
Citation is usually taught as a rule about honesty, which it is, but that framing makes it feel like a tax on writing. It is better understood as part of the argument. A citation tells your reader where a claim came from so they can check it, shows which parts of your paper are your contribution and which are inherited, and locates your work in a conversation. A paper with no citations is either unsupported or plagiarised. A paper with citations in the right places is one a reader can trust and build on.
Which style to use
You may use either APA 7th edition or Chicago author–date in my courses, unless a course outline says otherwise. Both are standard in political science: APA and APA-like formats dominate the North American journals, and Chicago author–date is at least as common in the discipline, particularly in comparative politics and in work with a historical dimension. Neither is better. Use the one your department, your other courses, or your own habits point you toward.
Chicago's other system — notes and bibliography, the footnote style — is standard in historically-oriented work and is fine where an assignment calls for it. It is more work to maintain by hand, so if you use it, use a reference manager.
What is not acceptable is mixing them. Half APA, half Chicago, with two entries in MLA because that is what the citation generator produced, is the most common formatting problem I see, and it is entirely avoidable. Consistency is the actual requirement here; the choice of style is not.
The differences that trip people up
The two styles carry the same information and differ mostly in punctuation. These are the points where students who switch between them go wrong:
| APA 7 | Chicago author–date | |
|---|---|---|
| Author and year | Comma between: (Jervis, 1976) | No comma: (Jervis 1976) |
| Page numbers | With "p." or "pp.": (Jervis, 1976, p. 58) | Bare number after a comma: (Jervis 1976, 58) |
| Two authors | Ampersand in parentheses: (Author & Author, 2020) | "and" spelled out: (Author and Author 2020) |
| Three or more authors | et al. from three onward | et al. from three onward |
| Author names in the list | Initials only: Jervis, R. | Full given names: Jervis, Robert. |
| Title capitalisation | Sentence case: Perception and misperception in international politics | Headline case: Perception and Misperception in International Politics |
| Article titles | No quotation marks | In quotation marks |
| Year placement | In parentheses after the author | After the author, no parentheses |
| Place of publication | Omitted | Omitted |
Two of those rows — the author rule and the place of publication — used to be genuine differences, and they are the reason a good deal of the advice you will find online is now wrong. The 18th edition of Chicago (2024) dropped the requirement to give a place of publication for books, and changed the in-text rule from "name all three authors" to "first author plus et al." from three onward — which is what APA already did. Library guides and citation generators built on the 17th edition still show the old forms. If something you are consulting tells you to write "Princeton, NJ: Princeton University Press" or "(Rathbun, Kertzer, and Paradis 2017)," it is following Chicago 17. Either is fine if you are consistent, but know which one you are using.
What needs a citation
Cite: direct quotations; paraphrases and summaries of someone else's ideas; statistics, data, and figures; specific facts that are not general knowledge; theories, frameworks, and concepts attributable to a person; images, tables, and charts you did not make; and anything you are disagreeing with.
You do not need to cite genuine common knowledge — that Canada is a federal state, that the Second World War ended in 1945. The workable test is whether an informed reader would accept the claim without wondering where it came from. When in doubt, cite. Over-citation is a stylistic flaw; under-citation is an academic offence.
Do not cite my lectures
This one surprises students, so here is the reasoning. A lecture is not a source. Nothing in it is original scholarship — it is my synthesis of other people's research, compressed and reorganised for teaching. When you cite "Paradis, lecture, 12 October," you are citing a middleman. What you actually want is the study the lecture was describing.
There is also a practical problem: your reader cannot retrieve a lecture. They cannot check what was said, in what context, or whether you understood it correctly. That is why APA treats unrecorded lectures as personal communication rather than as a source, and why Chicago keeps them out of the reference list. A citation that no one can follow is not doing the job a citation exists to do.
So if something from lecture is load-bearing in your paper, do one of these instead:
- Go to the source. Lecture slides cite their material; the readings are on the course outline. Find the study, read the relevant part, and cite that. You will almost always find the original says something more precise than my summary of it.
- Cite the scholarly source instead. Named concepts and frameworks — the security dilemma, the democratic peace, the selectorate model — need a citation to the work that developed them, not to the lecture where you first met them. A few things genuinely are common knowledge in the field and need no citation at all: that Canada is a parliamentary system, that the Cold War ended in 1991. The dividing line is whether the idea is attributable to someone.
- Ask me if you cannot find where something came from. Sometimes it is from a source that is not on the outline, and I will point you to it.
The same logic applies to lecture slides, course notes, textbook summaries, and recorded class sessions. It also applies to your own notes: what you wrote down is your paraphrase of my paraphrase of someone else's research, two steps removed from anything checkable.
Two narrow exceptions. If the assignment explicitly asks you to engage with course material as material — a reflection paper, for instance — then cite it as the assignment directs. And if you are analysing the lecture itself, which essentially never happens in my courses, it becomes an object of study and gets cited accordingly.
Quoting, paraphrasing, and the space in between
A quotation reproduces exact words in quotation marks with a page number. Use it when the wording itself matters — a definition, a famous formulation, a claim you are about to dispute. Most student papers quote far too much; in political science, quotations should be rare and short.
A paraphrase restates an idea in your own words and sentence structure, and still needs a citation with a page number. Changing a few words while keeping the author's sentence shape is not a paraphrase — it is what the academic integrity policies call patchwriting, and it is treated as plagiarism at both of my institutions even when a citation is present.
Original (constructed example)
"Democratic institutions constrain leaders not because voters punish bad foreign policy directly, but because opposition elites are freer to make its costs visible."
Patchwriting — not acceptable
Democratic institutions restrain leaders not because the electorate punishes poor foreign policy directly, but because opposition elites are more able to make its costs apparent (Smith, 2019, p. 44).
Genuine paraphrase
Smith (2019) locates the disciplining effect of democracy in elite competition rather than in mass opinion: what matters is that rivals within the political class have the freedom to publicise the costs of a policy, not that voters track foreign affairs closely (p. 44).
The reliable method: read the passage, close the book, write what it said from memory, then reopen and check that you have the idea right and the words different.
In-text citation basics
| Situation | APA 7 | Chicago author–date |
|---|---|---|
| Paraphrase, author not named in sentence | (Jervis, 1976) | (Jervis 1976) |
| Paraphrase, author named in sentence | Jervis (1976) argues that… | Jervis (1976) argues that… |
| Direct quotation | (Jervis, 1976, p. 58) | (Jervis 1976, 58) |
| Two authors | (Author & Author, 2020) | (Author and Author 2020) |
| Three or more authors | (Rathbun et al., 2017) | (Rathbun et al. 2017) |
| Organisation as author | (Statistics Canada [StatCan], 2024) | (Statistics Canada 2024) |
| Multiple sources at once | (Jervis, 1976; Rathbun et al., 2017) | (Jervis 1976; Rathbun et al. 2017) |
| Source quoted inside another source | (Waltz, 1979, as cited in Rathbun et al., 2017) | (Waltz 1979, quoted in Rathbun et al. 2017, S40) |
On that last one: cite the original wherever you possibly can. Secondary citation says to your reader that you did not go and look, and occasionally what you find when you do look is that the original does not say what the citing author claimed.
Reference list examples
These are the source types political science students get wrong most often. Each is given in both styles — use one or the other throughout, never both.
Two real works run through the examples: Jervis (1976), a single-authored book, and Rathbun, Kertzer, and Paradis (2017), a three-authored journal article. Everything else uses placeholder names and titles to show the shape of the format rather than to point at a real document.
Journal article
APA 7
Rathbun, B. C., Kertzer, J. D., & Paradis, M. (2017). Homo diplomaticus: Mixed-method evidence of variation in strategic rationality. International Organization, 71(S1), S33–S60. https://doi.org/10.1017/S0020818316000412
Chicago author–date
Rathbun, Brian C., Joshua D. Kertzer, and Mark Paradis. 2017. "Homo Diplomaticus: Mixed-Method Evidence of Variation in Strategic Rationality." International Organization 71 (S1): S33–S60. https://doi.org/10.1017/S0020818316000412.
Two things to notice. This article appeared in a special issue, so the issue number is S1 and the page numbers carry an S prefix — copy those exactly rather than tidying them. And both entries end with the DOI, which is preferred over a URL in both styles: it is permanent, and a journal URL frequently is not.
Book
APA 7
Jervis, R. (1976). Perception and misperception in international politics. Princeton University Press.
Chicago author–date
Jervis, Robert. 1976. Perception and Misperception in International Politics. Princeton University Press.
Chapter in an edited book
APA 7
Author, A. A. (2021). Title of the chapter. In B. B. Editor & C. C. Editor (Eds.), Title of the book (2nd ed., pp. 115–140). Publisher.
Chicago author–date
Author, Anna A. 2021. "Title of the Chapter." In Title of the Book, edited by Bob B. Editor and Carol C. Editor, 2nd ed., 115–40. Publisher.
Government document or report
APA 7
Global Affairs Canada. (2022). Canada's Indo-Pacific strategy. Government of Canada. https://www.international.gc.ca/…
Chicago author–date
Global Affairs Canada. 2022. Canada's Indo-Pacific Strategy. Government of Canada. https://www.international.gc.ca/….
Here the agency is the author and the government is the publisher, which is the usual arrangement for departmental documents. Where the author and the publisher really are the same body — a standalone organisation issuing its own report — APA has you name it once as author and omit the duplicate publisher line.
Parliamentary debate or committee evidence
APA 7
House of Commons Standing Committee on Foreign Affairs and International Development. (2024, March 12). Evidence (No. 87, 44th Parliament, 1st Session). Parliament of Canada. https://www.ourcommons.ca/…
Chicago author–date
House of Commons Standing Committee on Foreign Affairs and International Development. 2024. Evidence, no. 87, 44th Parliament, 1st Session, March 12, 2024. Parliament of Canada. https://www.ourcommons.ca/….
Treaty
North Atlantic Treaty, April 4, 1949, 34 U.N.T.S. 243.
Treaties, statutes, and cases follow legal citation conventions rather than either style's rules, so this form is the same whichever you are using. In Canada, the Canadian Guide to Uniform Legal Citation (the McGill Guide) is authoritative; for a political science paper, a clear and consistent form giving the instrument, date, and official source is sufficient.
Intergovernmental organisation report
APA 7
United Nations Development Programme. (2025). Human development report 2025. United Nations. https://hdr.undp.org/…
Chicago author–date
United Nations Development Programme. 2025. Human Development Report 2025. United Nations. https://hdr.undp.org/….
UN resolution
APA 7
UN Security Council. (2011, March 17). Resolution 1973 (S/RES/1973). United Nations. https://digitallibrary.un.org/…
Chicago author–date
UN Security Council. 2011. Resolution 1973. S/RES/1973, March 17, 2011. https://digitallibrary.un.org/….
Dataset
APA 7
Coppedge, M., Gerring, J., Knutsen, C. H., Lindberg, S. I., Teorell, J., & Altman, D. (2025). V-Dem country-year dataset v15 [Data set]. Varieties of Democracy Project. https://doi.org/…
Chicago author–date
Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Jan Teorell, and David Altman. 2025. "V-Dem Country-Year Dataset v15." Varieties of Democracy Project. https://doi.org/….
Cite the specific version you used, in either style. "V-Dem" without a version number is not reproducible, and reproducibility is the entire point of citing data — note too that v15 was superseded in 2026, so check which release you actually downloaded.
The author list above is abbreviated to save space: the official V-Dem citation names around thirty people. Large data projects publish the exact citation they want used, normally in the codebook or on the download page — use theirs rather than building your own, and follow your style's rule for long author lists. APA lists the first nineteen, then an ellipsis, then the last; Chicago lists up to six, and beyond that gives the first three plus et al.
Public opinion poll
APA 7
Angus Reid Institute. (2026, February 4). Title of the survey report [Survey report]. https://angusreid.org/…
Chicago author–date
Angus Reid Institute. 2026. "Title of the Survey Report." February 4, 2026. https://angusreid.org/….
In your text, give the fieldwork dates, sample size, and method as well — a poll cited without them cannot be evaluated by your reader.
Working paper or preprint
APA 7
Author, A. A. (2026). Title of the paper (NBER Working Paper No. 34567). National Bureau of Economic Research. https://doi.org/…
Chicago author–date
Author, Anna A. 2026. "Title of the Paper." NBER Working Paper 34567. National Bureau of Economic Research. https://doi.org/….
News article
APA 7
Reporter, R. (2026, January 18). Ottawa signals shift in Arctic defence posture. The Globe and Mail. https://www.theglobeandmail.com/…
Chicago author–date
Reporter, Robin. 2026. "Ottawa Signals Shift in Arctic Defence Posture." Globe and Mail, January 18, 2026. https://www.theglobeandmail.com/….
Software or R package
APA 7
Author, A. A., & Author, B. B. (2023). packagename: Title of the package (Version 2.0.0) [Computer software]. https://CRAN.R-project.org/package=packagename
Chicago author–date
Author, Anna A., and Bob B. Author. 2023. packagename: Title of the Package. R package version 2.0.0. https://CRAN.R-project.org/package=packagename.
Do not build these by hand. In R, citation("packagename") returns the citation the package authors want used, in full — including the complete author list, which for a widely-used package can run to two dozen names. Note also that APA's bracketed descriptor is [Computer software]; [R package] is not an APA descriptor, however sensible it looks.
Generative AI output
APA 7
Anthropic. (2026). Claude (Version [model and version you used]) [Large language model]. https://claude.ai
Chicago author–date
Anthropic. 2026. Claude (Version [model and version you used]). https://claude.ai.
Cite an AI output only where the output is itself your object of analysis — see the AI policy. Because your reader cannot retrieve the exchange, the prompt and the full response belong in an appendix whichever style you are using.
The two styles disagree here, so be careful whose advice you are following. APA gives AI output a full reference-list entry, as above, and specifically declines to treat it as personal communication on the grounds that the output does not come from a person. Chicago takes the opposite view: because the exchange is not retrievable by your reader, it prefers that you describe it in the text or in a note rather than list it in the references.
Errors I see every term
- Citing a source you did not open. Usually visible, because the claim attributed to it is not one it makes.
- A citation at the end of a paragraph covering four sentences from three sources. Cite at the point of use.
- No page numbers on quotations. Always required.
- Entries in the reference list that never appear in the text, or in-text citations missing from the reference list. Every entry must appear in both.
- Citation-generator output pasted in unchecked. These tools get capitalisation, editors, and journal names wrong constantly. Check what they produce.
- Citing the database rather than the source. JSTOR is not the publisher of the article.
- Broken or unstable URLs. Use the DOI when one exists; it is permanent, and a URL frequently is not.
- Self-plagiarism. Submitting work you wrote for another course is an offence at both institutions unless you have written permission from both instructors. Ask first — sometimes the answer is yes.
Tools
Use Zotero. It is free and open source, it has a browser connector that captures a citation from a journal page in one click, and it inserts formatted citations into Word, Google Docs, and LibreOffice while building your reference list as you go. Changing citation style becomes a two-second operation instead of an evening.
The caveat that applies to every reference manager: it is only as good as the metadata it captures. Check the entries, particularly for government documents, reports, and book chapters, where the automatic capture is least reliable.
For the authoritative rules, both libraries maintain style guides, and the APA Style site publishes its own examples free of charge.
Not sure how to cite something unusual — an archival document, a leaked cable, a dataset you built yourself? Ask me. The general principle behind every citation format is: give your reader enough information to find exactly what you used.