Skip to main content

Citing and referencing

What’s the difference between a citation and a reference?​

Citations are used within the text, while references are presented at the end. Citations are brief, usually mentioning only the author name and publication year. Their full details are given in the references, which are longer.

Example of a citation:

  • Yang (2020) explored how children perceived colors in books.

Example of a reference (a journal article in APA style):

Some citations may not be in the reference list, if there is no source to be referenced - for example, if you cite a personal source (‘personal communication’) or a data set that has been shared with you personally and is not published or indexed anywhere.

Can I cite a journal article that’s been retracted?​

In cases where you can leave a retracted paper out of your work, do leave it out. Papers are retracted for a reason, and shouldn’t be cited as if they are still published papers.

Only continue to cite a retracted paper if it is key to your work (for example, you challenge a theory introduced in it) and the reason for retraction doesn’t affect its validity within your work (for example, you do not use its results to support your case if these results were based on manipulated data).

As advised by Retraction Watch, you should acknowledge within the text that the work has been retracted, and cite both the original article and the retraction notice. The retraction notice is a separate, short publication that explains why a paper has been retracted. Look for guidelines on how to cite retracted papers and retraction notices; for example, these guidelines for APA.

What’s the difference between a bibliography, References, and a Works Cited list?​

Works Cited and References are the same thing: both are lists of references that you have cited in your work. Which name you use depends on your referencing style: MLA (Modern Language Association) uses ‘Works Cited’, and APA (American Psychological Association) uses ‘References’.

A bibliography is a list of all the sources you have used to prepare your work, including those you haven’t cited. This means a bibliography is usually longer than a Works Cited or References list. It is often numbered, while a Works Cited or References list is alphabetically sorted by the author’s last name. A bibliography may be required when using the Chicago style.

How do I know I’m not missing relevant references?​

To make sure you cover the most important references, you can:

  • Read a review paper on your topic. Reviews cover the most relevant literature on a topic, so you can quickly see which papers you should read.
  • Browse a few different databases, like Scopus, Web of Science, and PubMed (if you have access through your institution). You can also browse Google Scholar. This covers more material, but does need a quality check from you. Since Google indexes automatically, not all sources are of equally high quality.
  • Come up with a strong query to get relevant results from a database. For instance, if your paper is about economic growth and energy consumption, make sure your search includes not just these terms, but also other forms, synonyms, and related words. Also use wildcards and other syntax to make your query stronger. Find instructions on your database’s Help or Support pages.
  • Once you have a good query, rank papers by the number of citations to see which have been cited most - these should be key papers you can’t leave out. Remember, though, that a high number of citations doesn’t always mean a paper is good; citations can be outdated, or even negative mentions!

When do I need to add a citation and reference?​

You need to add a citation whenever you draw information that is not common knowledge from another source. For example, the results, conclusions, or opinions of other authors are things you should cite. A citation is needed when you paraphrase, quote, or otherwise refer to existing work - no matter the source (a journal article, book, data set, website, conference presentation, YouTube video, etc.).

Everything in your reference list should also be cited within your text (unless you’re using a bibliography). Citations in your text will usually also be in your reference list, but exceptions are citations of personal communications or unpublished materials, such as privately held data.

Find more tips about citing and referencing to avoid plagiarism here.

Should I add a reference if I give my opinion?​

This depends on whether your opinion follows those of others, or is new. For example, if you believe the performance of dentists should be evaluated differently and this has been expressed by many researchers, adding a few key sources will strengthen your point. For example: ‘The evaluation criteria for dentists should be reconsidered (Author 2020, Author 2021, Author 2022).’

If your opinion does not reflect a wider belief, it can still be useful to add references. Some might indirectly support your view; maybe the authors haven’t explicitly argued the same, but their results or conclusions support your argument. For example, if you write ‘The evaluation criteria for dentists should be reconsidered’, you could cite sources that have found that the current evaluation criteria aren’t solid.

Read more here.

How many references are enough for my essay, paper or thesis?​

It is often advised to use approximately one reference for each 150 words. However, this is just a rule of thumb. Keep in mind that:

  • How many references you need depends on the type of work and your message. If you’re writing a review paper to submit to a journal, you’ll include more references than if you’re writing an essay for your English class.
  • The number of references isn’t as important as the quality of these references and their relevance to your work. It’s quite easy to find a string of references to add after a statement, but a knowledgeable reader will quickly see whether these references are supportive and relevant.
  • Referencing norms differ between fields, authors, and journals. Take a look at essays, papers, or theses from your field to see how many citations are usually added to support statements.

Where should I put an in-text citation?​

You can put an in-text citation:

  • immediately after the author’s nameIf you want to emphasize the author of the source, place the citation right after the author’s name. For example:
    • Author (2021) explored the role of technology in first language development.
  • at the end of the sentence or clauseIf you want to emphasize the information and not the author, place the reference elsewhere in the sentence:
    • The role of technology in first language development has recently been explored (Author, 2021).
    • As technology plays an important role in first language development (Author, 2021), this study considers …

How can I check whether my reference list is complete?​

The easiest way to do this is to use a reference manager: a tool that helps you add, format, and quality-check your references and citations. Reference managers let you save your sources and cite them using a chosen referencing style (for example, APA or Harvard), and then automatically generate and update your reference list. This will ensure that whatever you cite also shows up in your reference list. Popular reference managers are Mendeley, Endnote, and Zotero.

Do I need to read all of the sources I cite?​

Yes, if you cite a source, you’re expected to have read at least the relevant parts of it. Do not cite sources that you haven’t at least skimmed, even if multiple sources cite it and you think you know what it’s about. The author(s) citing it may have misinterpreted it, or the paper may have more information that is relevant to your work.

If you come across a relevant citation in another work, but you cannot access the source yourself (for example because it is an old book that you can’t get hold of), you can use the ‘cited in’ principle. For example: (Sánchez, 1989, as cited in Ribeiro, 1992). Only use a citation from another source if you can’t read the original source yourself.

If you cite a work, it isn’t always necessary to read the whole text. If only a small part of it is relevant, and that part can be used independently, it’s fine to skip the rest. For example, if a paper lists examples of a phenomenon, you can cite these examples without reading the whole paper. However, in most cases, you’ll want to read the whole source, to make sure you understand the part you cite and that it is valid. For example, if you cite the conclusions of a study, you’ll need to read the entire work to make sure the conclusions are valid based on the study’s data, and that the conclusions also apply to your work.

How do I cite a source that is cited by another author?​

First, try to cite the primary source (the original, cited one) where you can. If this isn’t possible, you can cite a source as ‘cited in’ the source you read. How you do this depends on your referencing style. Here are examples for APA:

  • In-text citation: (Ribeira, 2009, as cited in Webb, 2012)
  • In your reference list: only list the secondary source (the one your read)

For most author-year styles, such as APA and Harvard, simply add the secondary source to your reference list (the source you read). For some numbered styles, add both the secondary and the primary source. If you’re not sure, check your referencing style handbook or ask your librarian for advice.

When is a reference too old to cite?​

This really depends on the discipline and the type of work. Citing ‘old’ publications is more common in the Arts and Humanities and in some Social Sciences than it is in most hard sciences such as Physics and Engineering. However, for all fields, citing older works makes sense when you are:

  • writing a review paper that discusses all key sources on a topic
  • explaining the origins or history of something
  • drawing from an influential publication
  • reusing an important quote

If these don’t apply, see if there are more recent sources you can cite. You can find tips here to ensure you’re not missing relevant studies.

How do I cite my own work?​

You cite your own published work as you would cite any other source, following your own referencing style. In most fields and text types, you will refer to yourself in the third person. For example:

  • Wiggins (2020) found that…

If it’s relevant for the reader to understand you were the author of a cited work, you can add this. This makes sense, for instance, when you’re running a follow-up study based on your previous work. This is how your self-citation note could look:

  • As suggested by the current author in a previous study (Wiggins, 2020), this follow-up experiment …

To cite research that hasn’t been published yet, in most referencing styles, you will keep the same referencing format but add a note to say the work is ‘under review’, ‘in press’, or otherwise. For example, in the APA style, the in-text citation would be ‘(Wiggins, in press)’ and the reference item would be ‘Wiggins, B. (in press) …’.

Can I cite Wikipedia? If so, how?​

In general, it’s best not to cite Wikipedia, as it isn’t a peer-reviewed scholarly source. If you find relevant information on Wikipedia, look for a scholarly source that covers this same information (maybe Wikipedia cites a source already). This can be a journal article, a report, a book chapter, a data set, or maybe a website.

This doesn’t mean Wikipedia isn’t useful for doing research. Wikipedia is great for:

I used someone else’s research data in my paper. Should I cite it?​

It’s good practice to do so, and if you’re looking to publish in a journal, you’ll find that many publishers encourage or even require it (see Springer Nature’s data policy as an example). This is because citing data gives credit to the researcher, and shows your reader what your work is based on.

How you cite data depends on the referencing style you follow as well as on the publisher and specific journal (if any) you submit to. In most cases, you will give a data set an in-text citation as well as an item in your reference list, just as you do with a book, chapter, or article. If a dataset is published in a repository, the repository might also give you citation guidelines (see for example these data sets with ready-made citations in Pangaea and Dryad).

Should I always include a DOI in my references?​

Yes, if a work has a DOI, it’s important to add this to the reference. A DOI is a unique identifier of a publication, which never changes. This means that even if the online location of a publication changes, the DOI will always point to it. Today, most scholarly online publications, including datasets and conference proceedings, have DOIs. If you add a DOI, there is no need to add a URL as well.

Is it necessary to cite when I paraphrase?​

Absolutely; no matter how much paraphrasing you do, if you use information from another source, you should cite it. It’s a misunderstanding that paraphrasing is a way to avoid plagiarism. Of course, you should paraphrase sources (unless you quote directly), but if you paraphrase without citing, you’re still plagiarizing.

Also see these sections on paraphrasing and plagiarism.

Should I use single or double quotation marks when citing an author? How do I quote within a quote?​

In US English, double quotation marks are used. Single quotation marks can then be used within the quotation, if necessary. For example: “The most frequently used verb by participants was ‘do’.”

In UK English, single quotation marks are preferred, with double quotation marks used within single quotation marks. So taking the above example: ‘The most frequently used verb by participants was “do”.’

Can I change the capitalization in direct quotes?​

The rules tend to differ between styles, publishers, and journal editors on this. But in general, you can change the capitalization in quotes as long as you indicate what letters (or words) you have changed. You can do this using squared brackets.

For example, to make a quote flow well within your sentence, you can change the first letter of the first word, and indicate this change with squared brackets:

  • Original sentence: ‘Most analyses are flawed, ….’
  • Changed capitalization: ‘Smith argues that “[m]ost analyses are flawed” (2018, p. 188).’