How to reference studies mentioned in a paper without having to read the original papers?

Question

I'm review the literature at the moment. Now I'm reading a paper that mention other studies. How can I reference them without reading the original papers?

Answer 1

Let us assume, for example, you read Doe (2011) and find Smith (1966) referenced therein. Technically, you can say something like "Smith (1966, cited in Doe, 2011)", or alternatively "(Smith 1966, cited in Doe, 2011)." The exact format depends on the format of the journal (it is also possible to phrase it "cited by" instead of "cited in").

That said, however, it is very dangerous to provide such quotes since you do not know if the person(s) citing the paper has understood it correctly. It is not unheard of that people cite for very odd reasons and not double-checking works cited may just propagate such errors.

So, it is possible but not recommended.

Answer 2

If it is relevant to your work, why wouldn't you read them? If they are not relevant, why would you cite them?

The only case I can think where it might make sense is if you are reading a review paper, and want to actually cite them as a collection rather than individually. Because there are a large number or for other reasons. Then you would write

Doe et al. collected in their recent reviews a large number of earlier work in (Doe, 2012 and references therein)

or

Doe et al. collected in their recent reviews a large number of earlier work in (Doe, 2012 and references 15–73 therein)

Answer 3

I wrote a post on writing literature reviews in psychology. Here's my advice:

Cited In: Good literature reviews do not use "Cited in". Literature reviews which summarise Author B’s citation of Author A’s work write: "as Author A (1999) says as cited in Author B (2002) …" . However, good literature reviews, when they see that Author B cites Author A, go and get Author A’s article, read it , and draw conclusions about it directly.

So it is only in rare occasions that you need to indicate that an article was cited by another author. Just because you learnt about a study because it was cited somewhere is generally not relevant. Read the original so that you know enough about it to incorporate it into your literature review.

Of course, there are many less common exceptions where you may wish to indicate the relationship between two papers:

• You want to discuss how Paper 1 uses Paper 2. For example, you might want to draw attention to how various papers have mis-used a citation in order to justify some misguided methodological practice.
• You are performing a meta-analysis and you want to indicate that you used a previous study to find references.

There is also a potential plagiarism issue around over-reliance on a single paper to generate your literature review. If for example, you took 95% of your references from the one paper, this would be questionable in general, but at least by using "cited in" you are being honest. Of course, I think this strategy of reviewing the literature should be avoided in general.

More generally, finding literature by following the citation trail backwards (by looking at references) and forwards (using tools like Google Scholar) forms part of a general set of strategies for finding literature.

Answer 4

As JeffE above said, "Don't". When you make assertions in your published work, they should be based on either

1. Something that you have demonstrated yourself in the work
2. Something that another person has demonstrated (so you cite them)
3. Something that is sufficiently well known to the intended audience that no citation should be necessary (you don't need to cite Newton for his law of gravitation for example)

To consider the bigger picture for a moment, something that is published does not suddenly become a part of scientific dogma, even if it's published in a top tier journal. The motto of the Royal Society is "Nullius in verba", that is, "take nobody's word for it". If you are relying on established results for your work, then reading these results is the very minimum you should do. In an ideal world where you would then proceed to replicate the results that you are depending on.

Of course, this is usually not practical because of constraints on your time but at the early stages of your studies you should at least consider replicating some results since:

1. It will give you a far greater familiarity with the domain
2. There's a very high chance that you will expose some misconception that you had held about the work you are citing
3. There's a reasonable chance that you will expose a deficiency in the original work

In closing, you should consider that every time you publish something, you have added to mankinds knowledge. Every time you cite a paper without reading it, you dilute that knowledge.